Search

Your search keyword '"Rajgopal A"' showing total 5,064 results
Start Over Author "Rajgopal A"
5,064 results on '"Rajgopal A"'

1. ClusterViG: Efficient Globally Aware Vision GNNs via Image Partitioning

Author

Parikh, Dhruv, Fein-Ashley, Jacob, Ye, Tian, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Distributed, Parallel, and Cluster Computing
Abstract

Convolutional Neural Networks (CNN) and Vision Transformers (ViT) have dominated the field of Computer Vision (CV). Graph Neural Networks (GNN) have performed remarkably well across diverse domains because they can represent complex relationships via unstructured graphs. However, the applicability of GNNs for visual tasks was unexplored till the introduction of Vision GNNs (ViG). Despite the success of ViGs, their performance is severely bottlenecked due to the expensive $k$-Nearest Neighbors ($k$-NN) based graph construction. Recent works addressing this bottleneck impose constraints on the flexibility of GNNs to build unstructured graphs, undermining their core advantage while introducing additional inefficiencies. To address these issues, in this paper, we propose a novel method called Dynamic Efficient Graph Convolution (DEGC) for designing efficient and globally aware ViGs. DEGC partitions the input image and constructs graphs in parallel for each partition, improving graph construction efficiency. Further, DEGC integrates local intra-graph and global inter-graph feature learning, enabling enhanced global context awareness. Using DEGC as a building block, we propose a novel CNN-GNN architecture, ClusterViG, for CV tasks. Extensive experiments indicate that ClusterViG reduces end-to-end inference latency for vision tasks by up to $5\times$ when compared against a suite of models such as ViG, ViHGNN, PVG, and GreedyViG, with a similar model parameter count. Additionally, ClusterViG reaches state-of-the-art performance on image classification, object detection, and instance segmentation tasks, demonstrating the effectiveness of the proposed globally aware learning strategy. Finally, input partitioning performed by DEGC enables ClusterViG to be trained efficiently on higher-resolution images, underscoring the scalability of our approach., Comment: Preprint

Published
2025

2. Towards Ideal Temporal Graph Neural Networks: Evaluations and Conclusions after 10,000 GPU Hours

Author

Yang, Yuxin, Zhou, Hongkuan, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning, Computer Science - Social and Information Networks
Abstract

Temporal Graph Neural Networks (TGNNs) have emerged as powerful tools for modeling dynamic interactions across various domains. The design space of TGNNs is notably complex, given the unique challenges in runtime efficiency and scalability raised by the evolving nature of temporal graphs. We contend that many of the existing works on TGNN modeling inadequately explore the design space, leading to suboptimal designs. Viewing TGNN models through a performance-focused lens often obstructs a deeper understanding of the advantages and disadvantages of each technique. Specifically, benchmarking efforts inherently evaluate models in their original designs and implementations, resulting in unclear accuracy comparisons and misleading runtime. To address these shortcomings, we propose a practical comparative evaluation framework that performs a design space search across well-known TGNN modules based on a unified, optimized code implementation. Using our framework, we make the first efforts towards addressing three critical questions in TGNN design, spending over 10,000 GPU hours: (1) investigating the efficiency of TGNN module designs, (2) analyzing how the effectiveness of these modules correlates with dataset patterns, and (3) exploring the interplay between multiple modules. Key outcomes of this directed investigative approach include demonstrating that the most recent neighbor sampling and attention aggregator outperform uniform neighbor sampling and MLP-Mixer aggregator; Assessing static node memory as an effective node memory alternative, and showing that the choice between static or dynamic node memory should be based on the repetition patterns in the dataset. Our in-depth analysis of the interplay between TGNN modules and dataset patterns should provide a deeper insight into TGNN performance along with potential research directions for designing more general and effective TGNNs.

Published
2024

3. Contextual Feedback Loops: Amplifying Deep Reasoning with Iterative Top-Down Feedback

Author

Fein-Ashley, Jacob, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning
Abstract

Conventional deep networks rely on one-way backpropagation that overlooks reconciling high-level predictions with lower-level representations. We propose \emph{Contextual Feedback Loops} (CFLs), a lightweight mechanism that re-injects top-down context into earlier layers for iterative refinement. Concretely, CFLs map the network's prediction to a compact \emph{context vector}, which is fused back into each layer via gating adapters. Unrolled over multiple feedback steps, CFLs unify feed-forward and feedback-driven inference, letting top-level outputs continually refine lower-level features. Despite minimal overhead, CFLs yield consistent gains on tasks including CIFAR-10, ImageNet-1k, SpeechCommands, and GLUE SST-2. Moreover, by a Banach Fixed Point argument under mild Lipschitz conditions, these updates converge stably. Overall, CFLs show that even modest top-down feedback can substantially improve deep models, aligning with cognitive theories of iterative perception.

Published
2024

4. Recognizing the Role of the Posterolateral Corner in Patients Undergoing Total Knee Arthroplasty for Fixed Varus Deformity

Author

Rajgopal A, Kumar S, and Aggarwal K

Subjects
varus deformity, medial release, posterolateral complex, constrained implants, Orthopedic surgery, RD701-811, Diseases of the musculoskeletal system, RC925-935
Abstract

Ashok Rajgopal, Sumit Kumar, Kalpana Aggarwal Institute of Musculoskeletal Disorders and Orthopaedics, Medanta- The Medicity Hospital, Gurugram, Haryana, 122001, IndiaCorrespondence: Ashok RajgopalInstitute of Musculoskeletal Disorders and Orthopaedics, Medanta- The Medicity Hospital, Gurugram, Haryana, 122001, IndiaTel +91 9811079211Email a_rajgopal@hotmail.comPurpose: Varus deformity is the commonest presentation of the arthritic knee requiring surgical intervention. While correctable deformities lend themselves to realignment options like unicompartmental replacement, fixed deformities often need a total knee replacement. Current treatment options for patients with fixed coronal varus malalignment undergoing total knee arthroplasty include varying degrees of medial soft tissue releases, often leading to infringement of the medial collateral ligament complex and increased use of constrained options. We describe the role of the posterolateral (PL) tether in a select subgroup of patients needing release to achieve correction and minimising use of constrained options.Patients and Methods: A total of 384 patients with fixed varus deformity were retrospectively evaluated and categorised on the basis of weight bearing x-rays into four groups, namely, knees with angulation (F1), angulation with subluxation and torsion (F2), medial translation (F3) and deformity with major medial bone loss (F4). From this cohort, we identified patients with a tight PL tether that needed release to achieve good correction. These were predominantly in the F2 and F3 subgroups. Functional scores and outcomes were evaluated at a mean follow-up of 120.23 months.Results: F1 cohort achieved good correction with medial soft tissue release, while F2 and F3 cohorts often needed a PL release. While functional outcomes and scores were comparable in both groups, survivorship was better in the group where release was done.Conclusion: We recognise the role of the PL tether in a subgroup of patients with recalcitrant fixed varus deformities. Sequential release helped achieve good outcomes with minimal use of constrained options.Level of Evidence: Three.Keywords: varus deformity, medial release, posterolateral complex, constrained implants

Published
2021

5. Effects of hospital facilities on patient outcomes after cancer surgery: an international, prospective, observational study

Author

Stephen R Knight, Catherine A Shaw, Riinu Pius, Thomas M Drake, Lisa Norman, Adesoji O Ademuyiwa, Adewale O Adisa, Maria Lorena Aguilera, Sara W Al-Saqqa, Ibrahim Al-Slaibi, Aneel Bhangu, Bruce M Biccard, Peter Brocklehurst, Ainhoa Costas-Chavarri, Kathryn Chu, Anna Dare, Muhammed Elhadi, Cameron J Fairfield, J Edward Fitzgerald, Dhruv Ghosh, James Glasbey, Mark I. van Berge Henegouwen, J.C. Allen Ingabire, T Peter Kingham, Marie Carmela Lapitan, Ismaïl Lawani, Bettina Lieske, Richard Lilford, Janet Martin, Kenneth A McLean, Rachel Moore, Dion Morton, Dmitri Nepogodiev, Faustin Ntirenganya, Francesco Pata, Thomas Pinkney, Ahmad Uzair Qureshi, Antonio Ramos-De la Medina, Aya Riad, Hosni Khairy Salem, Joana Simões, Richard Spence, Neil Smart, Stephen Tabiri, Hannah Thomas, Thomas G Weiser, Malcolm West, John Whitaker, Ewen M Harrison, Arben Gjata, Maria Marta Modolo, Sebastian King, Erick Chan, Sayeda Nazmun Nahar, Ade Waterman, Dominique Vervoort, Alemayehu Ginbo Bedada, Bernardo De Azevedo, Ana Gabriela Figueiredo, Manol Sokolov, Venerand Barendegere, Gerald Ekwen, Arnav Agarwal, Qinyang Liu, Juan Camilo Correa, Kalisya Luc Malemo, Jacques Bake, Jakov Mihanovic, Kamila Kunčarová, Julius Orhalmi, Hosni Salem, Jyri Teras, Aristotelis Kechagias, Alexis P Arnaud, Judith Lindert, Vasileios Kalles, Maria-Lorena Aguilera-Arevalo, Gustavo Recinos, Zsolt Baranyai, Basant Kumar, Harish Neelamraju Lakshmi, Sanoop Koshy Zachariah, Philip Alexander, Sunil Kumar Venkatappa, C Pramesh, Radhian Amandito, Christina Fleming, Luca Ansaloni, Gianluca Pellino, Ahmed M. Altibi, Ibrahim Nour, Intisar Hamdun, Ali M. Ghellai, Donatas Venskutonis, Tomas Poskus, Justas Zilinskas, Precious Malemia, Yong Yong Tew, Elaine Borg, Sarah Ellul, fatima Zahraa Wafqui, David W Borowski, Anne Sophie van Dalen, Cameron Wells, Harissou Adamou, Adesoji Ademuyiwa, Adewale Adisa, Kjetil Søreide, Sara Al Saqqa, Osaid Alser, Haya Tahboub, Helmut Alfredo Segovia Lohse, Sebastian Shu Yip, Piotr Major, António Sampaio Soares, Matei Razvan Bratu, Andrey Litvin, Armen Vardanyan, JC Allen Ingabire, Ahmad Gudal, Naif Albati, Jovan Juloski, Miran Rems, Sarah Rayne, Stephanie Van Straten, Yoshan Moodley, Irene Ortega Vázquez, Jaime Ruiz-Tovar, Kithsiri Janakantha Senanayake, Sujeewa Priyantha Bandara Thalgaspitiya, Omer Abdelbagi Omer, Anmar Homeida, Yucel Cengiz, Daniel Clerc, Muhammad Alshaar, Hanen Bouaziz, Yuksel Altinel, Matthew Doe, Maryna Freigofer, Ella Teasdale, Rakan Kabariti, Joshua Michael Clements, Stephen Richard Knight, Ahsan Ashfaq, Ijeoma Azodo, Gabriela Wagner, Ivan Trostchansky, Mayaba Maimbo, David Linyama, Helidon Nina, Amanda Zeko, Claudio Gabriel Fermani, Santiago Villalobos, Federico Carballo, Pablo Farina, Sebastian Guckenheimer, Marilla Dickfos, Ankit Ajmera, Chester Chong, Ralph Gourlay, Sikandar Hussaini, Yi Jia Lee, Adeeb Majid, Peter Martin, Rebecca Miles, Owen James Morris, Jamie Phua, William Ridley, Tarunpreet Saluja, Ryan Renxin Tan, Jen Teh, Anna Wells, Bharti Arora, Qaasim Dollie, Debbie Ho, Yanru Ma, Omattage Mahasha Perera, Anthony Truong, Amanda Caroline Dawson, Bryan Lim, Upuli Pahalawatta, Jacqueline Phan, Xiao-Ming Sarah Woon-Shoo-Tong, Andrea Yeoh, Lillian Charman, Andrew Drane, Sharon Laura, Charmaine Chu Wen Lo, Amy Mozes, Rita Poon, Hao Han Tan, Ellen Wall, Prakshi Chopra, Jasmine De Giovanni, Bal Dhital, Brian Draganic, Alexander Duller, Jonathan Gani, Yao Kuan Goh, Jun Young Jeong, Brendan McManus, Prakash Nagappan, Peter Pockney, Anya Rugendyke, Mahsa Sarrami, Stephen Smith, Vanessa Wills, Hsu Ven Wong, Geoffrey Ye, Geoffrey Zhang, Ethan Brooker, Daniel Feng, Bonnie Lau, Carlin Ngai, Sarah Birks, David Gyorki, Jaime Otero de Pablos, Ali Abbosh, Chris Gillespie, Ahmed Mahmoud, Bianca Kwan, Joshua Lawson, Andrea Warwick, Janne Bingham, Andrew J Cockbain, Nagendra Naidu Dudi-Venkata, Jordan Ellaby-Hall, Ben Finlay, Emily Humphries, Jade Pisaniello, Monique Pisaniello, Salma Salih, Tarik Sammour, Haidar Hadri Abd Wahab, April De Silva, Nicola Hayward, Kartik Iyer, Guy Maddern, Gian Andrea Prevost, Naga Annapureddy, Krishna Pranathi Settipalli, Jeremy Yeo, Lucy Hempenstall, Lily Pham, Shaun Purcell, Cherry Talavera, Ashish I Vaska, Gurpreet Chaggar, Phillip Chrapko, Annelise Cocco, Sarah Michelle Crystal Jade Coulter-Nile, Grahame Ctercteko, James French, Houchen Gong, Martijn Gosselink, Thuvarahan Jegathees, Ivan Jin, Michelle Kalachov, Kathryn Kiefhaber, Katherine Lee, Jason Luong, Steven Phan, Henry Pleass, Kelly Veale, Zhi Zeng, Angela Au, Ashe DeBiasio, Idy Deng, Jananee Myooran, Amrita Nair, Peter Stewart, Anton Stift, Lukas Walter Unger, Kerstin Wimmer, Nabila Ahmed, Syed Hasan, Saber Rahman, Margaret O'Shea, Greg Padmore, Adrian Peters, Pietro Perduca, Guenda Pulcina, Nicolas Tinton, Frederic Buxant, Elsa Dabin, Giulia Garofalo, Francis Dossou, Freddy Houehanou Rodrigue Gnangnon, Yacoubou Imorou Souaibou, Pako Motlaleselelo, Omphile Tlhomelang, Igor Lima Buarque, Gustavo Mendonça Ataíde Gomes, Aldo Vieira Barros, Ilia Batashki, Nikolai Damianov, Vladislav Stoyanov, Dragomir Dardanov, Svilen Maslyankov, Plamen Petkov, George Todorov, Evgeni Zhivkov, Aygulya Akisheva, Miguel Angel Castilla Moreno, Geno Genov, Ivelina Ilieva, Tsvetomir Ivanov, Martin Karamanliev, Azhar Khan, Emil Mitkov, Tsanko Yotsov, Boyko Atanasov, Nikolay Belev, Mihail Slavchev, Carlos Nsengiyumva, Elgan Jones, Simon Stock, Steve Kyota, James Brown, Tresor Mabanza K., Lemery Nigo Samuel, Chidi Otuneme, Ngwang Prosper, Franklin Umenze, Marylise Boutros, Natasha Caminsky, Sinziana Dumitra, Richard Garfinkle, Dominique Morency, Ebram Salama, Alexander Banks, Lorenzo Ferri, Haitian He, Amit Katz, Alexander Sender Liberman, Sarkis Meterissian, Allison Pang, Elena Parvez, Usmaan Hameed, Fahima Osman, Sangita Sequeira, Natalie Coburn, Alisha Jaffer, Paul Karanicolas, Matthew Mosseler, Reilly Musselman, Xinyuan Liu, Ching Wan Yip, Juan Sebastian Garces-Otero, Carolina Guzman, Sebastian Sierra, Andres Uribe Valencia, Paulo Andrés Cabrera Rivera, Saul Camelo, Andrea Gonzalez, Alejandro González-Orozco, Manuel Santiago Mosquera Paz, Carlos J- Perez Rivera, Felipe Gonzalez, Andres Isaza-Restrepo, Laura Nino- Torres, Natalia Arias Madrid, Maria Clara Mendoza Arango, Justin Tsandiraki, Damir Jemendžić, Branislav Kocman, Oliver Šuman, Renata Canic, Darko Jurišić, Ivana Karakas, Ana Krizanovic Rupcic, Vlatka Pitlovic, Josip Samardžić, Mario Kopljar, Ivan Bacic, Edgar Domini, Robert Karlo, Danijela Miljanić, Andrea Simic, Mariam Ahmed, Majdi Al Nassrallah, Rabiya Altaf, Talal Amjad, Ruba Eltoum, Heba Haidar, Alhassan Hassan, Omar Khalil, Marwan Qasem, Rommel Ramesh, Gautham Sajith, Maham Wisal, Jan Žatecký, Michele Bujda, Katerina Jirankova, Ales Paclik, Aya Abdallah, Mariam Abdulgawad Almogy, Esraa Ayman El-sawy, Ahmed Moustafa ElFayoumy, Nourhan Elghareeb, Nourhan Ahmed Esmat, Ahmed Fadel, Abdullah Habater, Heba Hamdy, Amr Hefni, Marwa Kamal, Norhan Mohamed Abobakr, Ahmed Sayed, Nancy Shaker, Ehab Taha, Hoda Tharwat, Omar Zakaria, Ibrahem Abdelmotaleb, Ali Al-Dhufri, Hamza S. Al-Himyari, Enas El sheikh, Asmaa Eldmaty, Aya Elkhalawy, Ahmed M.Elkhashen, Kithara Magdy, Safa Mostafa, Habib Doutoum Sadia, Mohamed mahmoud Saleh, Dina Samir, Mohamed Yahia Mohamed Ali, Mahmoud A. Nassar, Samar Abdelhady, Aly Abdelrazek, Israa Abdelsalam, Aya El-Sawy, Eman Essam, Mohamed Gadelkarim, Khaled Ghaly, Mohamed Hassabalnaby, Rana Masarani, Nourhan Mohamed Shaaban, Ahmed Sabry, Menatalla Salem, Nourhan Akram Soliman, Diaaaldin Zahran, Moustafa Ramadan Abou El.soud, Esraa Tarek Badr, Hala Borham, Nehal Elmeslemany, Mohammad Elsayed, Fawzia Elsherif, Sara Eslam, Gehad Gaber, Sondos Ibrahim, Yara Kamh, Abdelrahman Mahmoud, Shimaa gamal Mohamed, Eman Morshedy, Cinderella Omar, Fatima Salem Soliman, Shaza Abdelkawy, Naglaa Abdelmohsen, Mahmoud Abdelshakour, Ahmed Dahy, Norhan Gamal, Mohammed Gamal, Ahmad Hasan, Helal Hetta, Nehad Mousa, Mohamed Omar, Somia Rabie, Mahmoud Saad, Bakeer Saleh, Marwa Sayed Mohamed, Muhammad Shawqi, Heba Abdelhady Mousa, Mostafa Alnoury, Mohamed Elbealawy, Ahmed Elshafey, Muhammad Essam Ibrahim El Desouki Muhammad Ahmed, Mennatullah Ghonaim, Fawzy Hgag, Mohamed Ibrahim, Mahmoud Morsy, Mohamed Reda Loaloa, Ahmed Refaat, Hadeer Samir, Fatma Shahien, Mohamed Sobhy, Fathy Sroor, Esraa Abdellatif, Marina Adel, Amr Abdelghani Afifi, Eman Afifi, Marco Antaky, Amr Dawoud, Naira El Zoghby, Amira El-remaily, Ali Abdelazez Elzanfaly, Ahmed Gadallah, Fatma Alzahraa Gamal, Omar Hashem, Shrouk Medhat Youssef, Aliaa Muhammad Attyah, Malak Munir, Omar Shazly, Esraa Taha, Karim Wilson, Sawsan Adel, Asmaa Ali, Esraa Eid, Esraa Elhelow, Marwa Elmahdy, Bassant Elshatby, Amany Hossam el-din Zakaria, Ahmad Hossny, Eman Ibrahim, Ahmed M.Yonis, Maram Metwalli, Basant Yousry, Esraa Zid, Mina A Yacoub, Ahmed Abdelhakim, Nervana Abouelsoad, Mo'min Alkhatib, Ahmed Ashraf, Alaa Ashraf, Yasmin Elazab, Mahmoud Elfanty, Osama Elkabir, Mai Elsayed, Ahmed Elshimy, Hager Elsobky, John Eskander, Ahmed Gad, Ward Hamsho, Noura Khaled Abdelwahed, Menna Magdy, Dalia Moharam, Abeer Osama, Shereen Ramadan, Radwa Roum, Taqwa Sayed, Tarneem Shehada, Ahmed Mohy Zidan, Khalid Abbas, Amr Ali, Mohamed Attia, Mohamed Balata, Ayman El Nakeeb, Mohamed Ibrahim Elsayed Elewaily, Ahmed Elfallal, Hossam Elfeki, Ahmed Elkhadragy, Sameh Emile, Helmy Ezzat, Hasnaa Hosni, Islam Mansour, Waleed Omar, Gehad Othman, Kareem Sadek, Mostafa Shalaby, Noura Shehab-Eldeen, Rawda Anas khalifa, Helmy Badr, Mostafa Eldeep, Ahmed Eldeep, Amany Eldoseuky mohammed, Salwa Khallaf, Eman Magdy Hegazy, Rokia Mahmoud, Pola Mikhail, Mahmoud Morsi, Sara Mowafy, Dina Raafat, Amina Safy, Marwa Sera, Ahmed shible Sera, Mostafa Salim Mohamed AbdAllah, Muhammad Abdelkader, Abdulrahman Osama Abdou, Ahmedgaber Ahmed, Shireen Gaafar, Fatma Ibrahim negm, Mina Lapic, Ahmed Maher, Hagar Mahmoud, Ahmed Mostafa, Mohamed Samir, Fatma Samy, Nourhan Semeda, Hind I. Shalaby, Alaa El-taweel, Ahmed Galal Elnagar, Ahmed Gamal Hemidan, Mohamed Hussein, Ahmed.A. Kandil, Mf Moawad, Ayat Allah Nasser Hamamah, Mostafa Soliman, Mohamed Abdelkhalek, Noura Abdelmaksoud Tawakel, Ahmed Mohamed Abdelwahed, Alrawy Abdou, Khalid Atallah, Mohammed Yasser Elsherbeny, Eman Emara, Mohamed Hamdy, Omar Hamdy, Amira Haron, Salma Ismail, Islam Hany Metwally, Nihal Mohamed Hamed Elgaml, Ahmed Nassar, Basel Refky, Mirna Sadek, Mahmoud Saleh, Asmaa Yunes, Mai Zakaria, Mohammed Zuhdy, Notila Fayed, Mohammed Mustafa Hassan Mohammed, Sander Kütner, Priit Melnik, Indrek Seire, Toomas Ümarik, Eppu Ainoa, Verner Eerola, Hanna Koppatz, Laura Koskenvuo, Ville Sallinen, Sini Takala, Jevgeni Katunin, Arto Turunen, Niki Christou, Muriel Mathonnet, Vincent Lavoue, Krystel Nyangoh Timoh, Lucie Soulabaille, Romain Lesourd, Aude Merdrignac, Laurent Sulpice, Benoît André, Elodie Chantalat, Charlotte Vaysse, Bertrand Dousset, Sebastien Gaujoux, Gregory Martin, Octavian Clonda, Domantas Juodis, Klaus Kienle, Andras Mravik, Samuel Palmer, Gabor Szabadhegyi, Anita Eseenam Agbeko, Solomon Gyabaah, Frank Enoch Gyamfi, Nuhu Naabo, Atta Owusu senior, Joseph Yorke, Frank Owusu, Francis Abantanga, Theophilus Teddy Kojo Anyomih, Abdul-Jalilu Mohammed Muntaka, Emmanuel Owusu Abem, Mohammed Sheriff, Paul M. Wondoh, Dimitrios Balalis, Dimitrios Korkolis, Georgios Gkiokas, Eirini Pantiora, Theodosios Theodosopoulos, Argyrios Ioannidis, Konstantinos Konstantinidis, Sofia Konstantinidou, Nikolaos Machairas, Anna Paspala, Anastasia Prodromidou, Christos Chouliaras, Konstantinos Papadopoulos, Ioannis Baloyiannis, Ioannis Mamaloudis, George Tzovaras, Ioanna Akrida, Maria-Ioanna Argentou, Stylianos Germanos, Evangelos Iliopoulos, Ioannis Maroulis, George Skroubis, George Theofanis, Christos Chatzakis, Orestis Ioannidis, Lydia Loutzidou, Panagiotis Karathanasis, Nikolaos Michalopoulos, Charalampos Theodoropoulos, Dimitrios Theodorou, Tania Triantafyllou, Zoe Garoufalia, Natasha Hasemaki, Michalis Kontos, Gregory Kouraklis, Stylianos Kykalos, Theodore Liakakos, Eustratia Mpaili, Alexandros Papalampros, Dimitrios Schizas, Athanasios Syllaios, Ekaterini Christina Tampaki, Antonios Tsimpoukelis, Maria Ioanna Antonopoulou, Eirini Deskou, Dimitrios K. Manatakis, Dimitrios Papageorgiou, Menelaos Zoulamoglou, Christos Anthoulakis, Michalis Margaritis, Nikolaos Nikoloudis, Veronica Campo, André Ceballos, Mario-Andrés Flores, Waleska Giron, Donghyun Ko, Gabriel Martinez, Verónica Rivera Lara, Nataly Rueda, Andres Sanchez, Jorge Carlos Guillermo Tejeda Garrido, Alvaro Eduardo Alvarez Rivera, Elvis Benjamin Bamaca Ixcajoc, Lilian Elizabeth Barreda Zelaya, Patricia Chacòn-Herrera, Ligia Margarita Corea Ruiz, Guillermo Echeverria-Davila, Mario Garcia, Danilo García, Edgar Fernando Gutiérrez Mayen, Noriega José, Nery Mazariegos, Diego Méndez, Michael Paniagua Espinoza, David Bardos, Marton Benke, Kristof Illes, Balint András Kokas, Réka Szabó, Akhila Appukuttan, Anjitha Asok, Vijaykumar D.k, Kapil Malik, Praveen Ravishankaran, Ritesh Tapkire, Guru Moorthy, Joyner Abraham, Ramesh Muthuvel, John Alapatt, Abhay Kattepur, Nizamudheen Pareekutty, Mebanshanbor Garod, Caleb Harris, Cliff Wanniang, Ashish Gupta, Deepak Nehra, Sanjeev Parshad, Rajgopal Acharya, Rajendra Badwe, Manish Bhandare, Urvashi Jain, Karishma Kirti, Nita Nair, Shailesh Shrikhande, Purvi Thakkar, Premkumar Anandan, Archana C S, Arun Holenarasipur Narasannaiah, Tejaswi Jagarlamudi, Rashmi M R, Mallikarjuna Manangi, Abhishek Raghavendra, K. Seshagiri Rao, Vinay S, Vinay Sajjan, Aneesh Shenoy, Santhosh Shivashankar Chikkanayakanahalli, Kavya Tharanath, Sushmita V, Peter Adidharma, Raksheeth Agarwal, Phebe Anggita Gultom, Ghafur Rasyid Arifin, Matthew Billy, Zatira Elfizri, Alessa Fahira, Devi Felicia, Triana Hardianti Gunardi, Nadya Johanna, Nadia Rahmadiani Nugrahadi, Sonar Soni Panigoro, Siti Rahmayanti, Retta Catherina Sihotang, Santi Yuanita Brata, Hadi Winoto, Nastaran Barati, Manoochehr Karami, Hamidreza Khorshidi, Homa Naderifar, Mazin A. Abdulla, Maggie Coleman, Ronan J Doherty, Rob Hannon, Brenda Murphy, Aine Stakelum, Des Winter, Lylas Aljohmani, Richard Farnan, Yeshey Seldon, Tanna Tan, Shriya Varghese, Mohammad Alherz, Muaaz Ather, Mohammad Bajilan, Vivien Graziadei, Isobel Pilkington, Omar Quidwai, Paul Ridgway, Haaris Shiwani, Abd al-Rahman Tahir, Eimear Blunnie, Daniel Burke, Niall Kennedy, Kate Macdonagh, Maeve O'Neill, Siobhan Rooney, Giuseppe Falco, Guglielmo Ferrari, Simone Mele, Gabriela Elisa Nita, Lara Ugoletti, Maurizio Zizzo, Gianmaria Confalonieri, Giovanni Pesenti, Fulvio Tagliabue, Gianluca Baronio, Deborah Ongaro, Giacomo Pata, Bruno Compagnoni, Renato Salvadori, Lucio Taglietti, Nicola D'Alessandro, Pierpaolo Di Lascio, Giovanni Pascale, Luca Bortolasi, Tommaso Campagnaro, Massimo Carlini, Giorgio Lisi, Davide Lombardi, Corrado Pedrazzani, Domenico Spoletini, Giulia Turri, Paola Violi, Donato Francesco Altomare, Fabrizio Aquilino, Nicola Musa, Vincenzo Papagni, Arcangelo Picciariello, Leonardo Vincenti, Dario Andreotti, Savino Occhionorelli, Matteo Tondo, Stefano Maria Massimiliano. Basso, Paolo Ubiali, Riccardo Cirelli, Marco Enrico Mario Maino, Guglielmo Niccolò Piozzi, Emanuele Picone, Rosa Scaramuzzo, Giovanni Sinibaldi, Alfonso Amendola, Lorenzo Anastasio, Luigi Bucci, Emanuele Caruso, Antonio Castaldi, Sara Di Maso, Vincenza Paola Dinuzzi, Giovanni Esposito, Maria Gaudiello, Mariano Cesare Giglio, Paola Antonella Greco, Gaetano Luglio, Andrea Manfreda, Ester Marra, Federica Mastella, Gianluca Pagano, Roberto Peltrini, Vincenzo Pepe, Michele Sacco, Viviana Sollazzo, Giovanni Spiezio, Ettore Cianchetti, Nunzia Menduni, Michele Maria Carvello, Francesca Di Candido, Antonino Spinelli, Fabio Corsi, Luca Sorrentino, Fabio Marino, Emanuele Luigi Giuseppe Asti, Luigi Bonavina, Emanuele Rausa, Martina Asta, Andrea Belli, Francesco Bianco, Carmela Cervone, Paolo Delrio, Armando Falato, Andrea Fares Bucci, Rita Guarino, Ugo Pace, Daniela Rega, Emilia De Luca, Gaetano Gallo, Giuseppe Sammarco, Giuseppe Sena, Giuseppina Vescio, Letizia Santandrea, Giampaolo Ugolini, Davide Zattoni, Nicola Chetta, Gaetano Logrieco, Serafino Vanella, Gianluca Garulli, Nicola Zanini, Andrea Bondurri, Francesco Cammarata, Francesco Colombo, Diego Foschi, Giulia Maria Beatrice Lamperti, Anna Maffioli, Gianluca Matteo Sampietro, Al'ona Yakushkina, Gloria Zaffaroni, Enrico Cicuttin, Maria Grazia Sibilla, Harmony Impellizzeri, Marco Inama, Gianluigi Moretto, Sylvie Mochet, Elisa Ponte, Antonella Usai, Stefano Mancini, Andrea Sagnotta, Luigi Solinas, Elisa Bolzonaro, Nicolò Tamini, Gianluca Curletti, Raffaele Galleano, Michele Malerba, Sofia Campanella, Gianfranco Cocorullo, Francesco Colli, Paolino De Marco, Nicolò Falco, Tommaso Fontana, Leonel jospin Kamdem Mambou, Antonella La Brocca, Leo Licari, Brenda Randisi, Giovanna Rizzo, Giulia Rotolo, Giuseppe Salamone, Roberta Tutino, Paolina Venturelli, Stefano Malabarba, Alessandro Sgrò, Ivan Vella, Bruno Cirillo, Daniele Crocetti, Giorgio De Toma, Pierfrancesco Lapolla, Andrea Mingoli, Paolo Sapienza, Angela Belvedere, Stefania Bianchini, Margherita Binetti, Arianna Birindelli, Valeria Tonini, Mauro Podda, Fabio Pulighe, Michele De Rosa, Lorenzo Bono, Felice Borghi, Paolo Geretto, Maria Carmela Giuffrida, Corrado Lauro, Alessandra Marano, Luca Pellegrino, Paola Salusso, Diego Sasia, Michela Campanelli, Alberto Realis Luc, Mario Trompetto, Roberto Cardia, Nicola Cillara, Antonio Nicola Giordano, Antonio Costanzo, Mario Alessandro Giovilli, Luca Turati, Silvestro Canonico, Guido Sciaudone, Francesco Selvaggi, Lucio Selvaggi, Nader Albsoul, Ahmad AlBsoul, Ala'a Aldeen Alkhatib, Osama Alsallaq, Justin Z. Amarin, Rami Ayoub, Isam Bsisu, M S El Muhtaseb, Mohammad Jabaiti, Jamal Melhem, Yasmeen Z. Qwaider, Mohammad Hasan Salameh, Ahmad Suleihat, Haya H. Suradi, Mohammad Alammarin, Almoutuz Aljaafreh, Mohammad Bani hani, Zeina Bani hani, Farah Bani Hani, Toqa Fahmawee, Shadi Hamouri, Cyrine Katanani, Ra'fat Tawalbeh, Tamara Tawalbeh, Hassan Zawahrah, Mohamad K. Abou Chaar, Lana Abusalem, Mahmoud Al-Masri, Hani Al-Najjar, Lutfi Barghuthi, Zahra Ahmed, Adnan Maulana, Omar Ngotho, Charbel Kamau, Aruyaru Stanley Mwenda, Fridah Bosire, Elizabeth Mwachiro, Robert Parker, Ian Simel, Kimutai Sylvester, Abdulmunem Ahmed Mustafa Althini, Sofian Elbarouni, Aya Elseed Elbeshina, Ahmed Gwea, Ans Malek, Wedad Albashir Masoud Farag, Abdulwahab Abdalei, Abu Baker Abdel Malik, Areej Abo-khammash, Ma'aly Abuhlaiga, Nour Adnan, Marwa Albaggar, Asma Alfitory, Asma Aljanfi, Fakhruddin Almuzghi, Zohoor Altumei, Fatima Alzabti, Hana Ashoushan, Mohamed Assalhi, Joma Azzubia, Sondos Bnhameida, Malik Delhen, Houssein Elshafei, Hana Elteir, Fatima Esbaga, Abdel Aziz Gobbi, Fatma Hamouda, Hamdan Hilan, Rania Ismail, Fieruz Jebran, Muataz Kasbour, Galia Maderi, Saja Mohammad, Burooj Mohammed, Habib Murtadi, Hamassat Mustafa, Mohamed Rajab, Sarah Trenba, Mariam Wafaa, Eman Al Sagheir, Alabas Almigheerbi, Ahmed Alzahaf, Sumayyah Ghayth Bahroun, Najah Ben Dallah, Mahmoud Elshaibani, Haitem Eswaye, Maha Karar, Samah Omar, Eman Younes, Maha Younes, Dafer Zreeg, Saleh Abujamra, Firas Ashour, Mala Elgammudi, Wesal Omar F. Aljadidi, Enas Saddouh, Randa Sharif, Aya Alabuzidi, AbdulMawlay Alwerfally, Sarra Aribi, Fatma Bibas, Taha Elfaituri, Yasmine Elhajjaji, Ala Khaled, Wegdan Khalil, Tesneem Layas, Enas Soula, Ahmed Tarek, Muad fathi khalleefah Abu hallalah, Hazem Abdelkarem Ahmed, Tagwa Alsharef, Abdulsalam Ali Ben Saoud, Tasnim El Gharmoul, Ahmed Elhadi, Safa Elrais, Abdulhalim Shebani, Heba Zarti, Asaid Zeiton, Marijus Ambrazevicius, Nerijus Kaselis, Migle Stakyte, Oleg Aliosin, Agne Cizauskaite, Sarunas Dailidenas, Vitalijus Eismontas, Migle Kybransiene, Vitalija Nutautiene, Narimantas Samalavicius, Dainius Simcikas, Algirdas Slepavicius, Albinas Tamosiunas, Nerijus Ubartas, Paulius Zeromskas, Saulius Bradulskis, Edvinas Dainius, Juozas Juočas, Egle Kubiliute, Juozas Kutkevičius, Aurimas Opolskis, Audrius Parseliunas, Andrejus Subocius, Egle Virbickaite, Diana Zuikyte, Algirdas Bogusevicius, Kristina Buzaite, Daiva Čepuliené, Ieva Cesleviciene, Vaidotas Cesna, Jolanta Gribauskaite, Povilas Ignatavicius, Mantas Jokubauskas, Monika Liugailaitè, Ernest Margelis, Ruta Mazelyte, Lina Pankratjevaitè, Matas Pažusis, Agne Rackeviciute, Justina Saladyte, Monika Škimelytè, Vygintas Šlenfuktas, Monika Sudeikyte, Algimantas Tamelis, Tomas Vanagas, Žygimantas Žumbakys, Aivaras Atkociunas, Audrius Dulskas, Justas Kuliavas, Justas Birutis, Sigitas Paškevičius, Mindaugas Šatkauskas, Donatas Danys, Matas Jakubauskas, Lina Jakubauskiene, Marius Kryzauskas, Vytautas Lipnickas, Gabija Makūnaitè, Fanjandrainy Rasoaherinomenjanahary, Herizo Rasolofonarivo, Luc Hervé Samison, Bitiel Banda, Vanessa Msosa, Ahmad Imran Ahmad Izzuddin, Andre Das, Ying Yee Gan, Tan Shong Sheng, Jia yng Siaw, Mohd Fadliyazid Ab Rahim, Dyg Zahratul Hamrak Abang Jamari, Nurfariza Che Husin, Muhd Yusairi Kamarulzaman, Yi Ping Lim, Nil Amri Mohamed Kamil, Mohd Razeen Mohd Hassan, Saidah Mohd Sahid, Johari Mustafa, Elaine Hui Been Ng, Wan Khamizar Wan Khazim, Ng Chang Ern, P.g. Lingeshan, Syariz Ezuan Sulaiman, Sue Ean Ang, Muhammad Navid Bin Mohamad Sithik, Yih Jeng Cheong, Mahadevan Deva Tata, Law Jia Xian, Aravinthan Kadravello, I-Ern Koh, Li-Yen Ng, Yuki Julius Ng We Yong, Kandasami Palayan, Chi Xuan Sam, Phuah Siow Jin, Jeremy Tan Ern Hwei, Yita Tang, Alvin Zubin Ter, Michael Pak-Kai Wong, Andee Dzulkarnaen Zakaria, Zaidi Zakaria, Fitjerald Henry, Thyivya Kalaiselvan, Muhammad Fairuz Shah Abd Karim, Mohamed Rezal Abdul Aziz, Nora Abdul Aziz, Tak Loon Khong, Peng Choong Lau, Hiong Chin Lim, April Camilla Roslani, Jonathan Chen Ken Seak, Sui-Weng Wong, Lai Fen Wong, Leow Yeen Chin, Mercy Chinemerem Anyanwu, Zachary Busuttil, Thomas Calleja, Kurt Lee Chircop, Ruth Cutajar, Andrew Michael Dimech, Joseph Galea, Kiara Gascon Perai, Ruth Gatt, Lisa Kelman, Elizabeth Micallef, Favour Nwolu, Kim Sammut, Joanna Thompson, Sean Warwicker, Matthew Zammit, Fernando Cordera, Efraín Cruz González, Jorge Sánchez-García, Francisco José Barbosa Camacho, Francisco Javier Barrera López, Carlos Jose Zuloaga Fernandez del Valle, Eric Acosta, Iván Romarico González Espinoza, Perla Moreno, Ana Olivia Cortes-Flores, Clotilde Fuentes Orozco, Alejandro Gonzalez Ojeda, Samantha Corro Díaz González, Laura Martinez, Bonifacio Mosqueda Amador, Armando Novoa, Dennet Arturo Olazo Espejo, Alejandro Jimenez, Federico Lopez Rosales, Elva Gabriela Vanoye, Luis Alberto Garcia Gonzalez, Roberto Carlos Miranda-Ackerman, Manuel Solano-Genesta, Alethia Alvarez-Cano, Hector Hugo Romero-Garza, Heriberto Medina-Franco, Lorelí Mejía-Fernández, Noel Salgado-Nesme, Omar Vergara-Fernandez, Guadalupe Montserrat Gutiérrez-Mota, Francisco Xavier Hernandez Vera, Anabella Llantada Lopez, Gilberto Morgan Villela, Felipe de Jesus Ramirez Padilla, Walezka Tapia Marin, Mónica Martínez Maldonado, Ramses Sánchez Suárez, José Manuel Troche, Chaymae Benyaiche, Oumaima Outani, Souadka Amine, Amine Benkabbou, Anass Mohammed Majbar, Raouf Mohsine, Ali Rafik, Thida Oung, Moe Moe Tin, Philipp Plarre, Anna Alberga, Nina Sluiter, Jurriaan Tuynman, Robin Blok, Didem Cömert, Roel Hompes, Marianne Kalff, Merel Elisabeth Stellingwerf, Pieter Tanis, Mark van Berge Henegouwen, Elise Maria van Praag, Daan Wisselink, Michael Gerhards, Josephine Lopes Cardozo, Emma Westerduin, Joske de Jonge, Aaw van Geloven, Kaz van Schilt, Frank den Boer, Simone Stoots, Stijn Vlek, Jamie Adams, Ibrahim S. Al-Busaidi, Gabrielle Budd, Seung il Choi, Michael Jen Jie Chu, Anurag Ganugapati, Lucy McKinstry, Rebecca Pascoe, Simon Richards, Kenrick Rosser, Annie Stevenson, Rebecca White, Shebani Farik, Jin Kwun, Ahmed Murad, Sarah Cowan, Timothy Hall, Michael Hayton, Laminou Malam Sani, Souleymane Oumarou Garba, Ibrahim Amadou Magagi, Oumarou Habou, Halima Aliyu, Muhammad Daniyan, Tunde T. Sholadoye, Lawal Abdullahi, Lofty-John Anyanwu, Aminu Mohammad Mohammad, Abubakar Bala Muhammad, Abdurrahman Abba Sheshe, Ibrahim Suleiman, Alaba Adesina, Ajibola Awolowo, Clement Onuoha, Omotayo Salami, Ogechukwu Taiwo, Agboola Taiwo, Stephen Kache, Jerry Godfrey Makama, Danjuma Sale, Olajide Abiola, Akinlabi Ajao, Anthony Ajiboye, Amarachukwu Etonyeaku, Julius Olaogun, Ademola Adebanjo, Opeoluwa Adesanya, Michael Olatunji Afolayan, Olanrewaju Balogun, Ayomide Makanjuola, Samuel Nwokocha, Rufus Wale Ojewola, Thomas Olagboyega Olajide, Adewale Aderounmu, Abdul-Rashid Adesunkanmi, Augustine Agbakwuru, Adeleke Akeem Aderogba, Olusegun Isaac Alatise, Olukayode Arowolo, Oladejo Lawal, Tajudeen Mohammed, Chinedu Ndegbu, Olalekan Olasehinde, Funmilola Wuraola, Akinbolaji Akinkuolie, Arinzechukwu Mosanya, Omobolaji Ayandipo, Peter Elemile, Taiwo Akeem Lawal, Samuel Ali SANI, Stephen Garba, Rebecca Hauwa SANI, Samson Olori, Henry Onyebuashi, Ifeanyi Umoke, Adedire Adenuga, Ademola Adeyeye, Olufemi Habeeb, Bashir Lawal, Abdulrasheed Nasir, Eirik Kjus Aahlin, Didrik Kjønås, Elisabeth Myrseth, Jibran Abbasy, Abdul Alvi, Omair Saleem, Asma Afzal, Anam Nazir, Muhammad Farooq, Ayesha Liaqat, Syed Asghar Naqi, Ali Raza, Muzna Sarfraz, Muhammad Sarwar, Muntaha Banglani, Ambreen Munir, Rahmat Sehrish, Bushra Ayub, Raza Sayyed, Amna Altaf, Saima Ayub, Komal Saeed, Bilal Syed, Sana Amir Akbar, Abdul Wahid Anwer, Ruqayya Naheed Khan, Amina Iqbal Khan, Shahid Khattak, Sameen Mohtasham, Muhammad Asad Parvaiz, Aamir Ali Syed, Abdul Basit Ansari, Noman Shahzad, Tanwir Khaliq, Isbah Rashid, Shahzad Hussain Waqar, Hasan Abu Al-saleem, Amjad Abu Alqumboz, Mohammad Alqadi, Adham Amro, Rawan Assa, Eman Awesat, Rawan Ayyad, Mohammed Hammad, Ayat Haymony, Bassel Hijazi, Bara Hmeidat, Rowaa Lahaseh, Aseel Qawasmi, Alaa Rajabi, Mohammed Shehada, Sundus Shkokani, Yasmine Yaghi, Nadine Yaghi, Mohammad AlZohour, Mohammad Farid, Yousef Mahmoud Habes, Wesam Juba, Yanal Nubani, Abdelrahman Rabee, Mohammad Sa'deh, Saeed Abed, Iyad Al basos, Mohammad Alswerki, Dina Ashour, Israa Awad, Samar Diab, Alaa El Jamassi, Sahar El-Kahlout, Somaya Elhout, Ahmed N K Hajjaj, Doaa Hasanain, Baraa Nabil hajjaj, Mohammed Obaid, Eman Saikaly, Ahmed Salhi, Hiba Al-Tammam, Murad Almasri, Muath Baniowda, Doha Beshtawi, Ali Horoub, Rami Misk, Bayan Mohammad, Rami Qasrawi, Tasnim Sholi, Samar Abu-Nimeh, Abrar Abu-srour, Sadi A. Abukhalaf, Samer Adawi, Barah Alsalameh, Kholoud Ayesh, Muawiyah Elqadi, Ahmad Hammouri, Fatima Karim Mustafa, Natalie Marzouqa, Shatha Melhem, Dima Miqdad, Balqees Mohamad, Mhammed Rawhi, Ayman B. Abu Ahammala, Ahmed Abu Ataya, Israa Abu Jayyab, Samar Al-Shwaikh, Othman Alagha, Mohammed Alasttal, Haneen Awadallah, Mahmood Elblbessy, Jehad Fares, Akram Jarbou, Ibtisam Mahfouz, Moath A. Albahnasawi, Asmaa' Abo mahadi, Hasan Abuelhatal, Ayham Abuelqomboz, Abdelrahman Almoqayyad, Abdallah Alwali, Reem Balaawi, Mahmoud Hamouda, Mohammed Humeid, Abdullah Jedyan, Tasneem Mahmoud Abu hamam, Ghadeer Matar, Ali Salem, Tahani Samra, Nureddin Shaheen, Karam Shihada, Ayoob A.Nemer, Mahmoud Abu Al Amrain, Abdulwhhab Abu Alamrain, Najlaa Abu Jamie, Mohammed R. Abu-Rous, Nada Alfarra, Mohammed AlTaweel, Noor Alwhaidi, Ramadan Hamed, Bader Saqqa, Ahmad Shaheen, Dana Aljaber, Loay Aljaberi, Malak Alwaheidi, Assef Jawaada, Hani Khaldi, Rami Qahoush, Jalil Qari, Rana Saadeh, Ahlam Salim, Aseel Yacoub, Abbas Abbas, Rana Abu shua'ib, Baraa Abu Zainah, Mahmoud AbuSirrees, Basheer Babaa, Ola Barhoush, Asef Belal qadomi, Laith Daraghmeh, Reema Haji, Alaa Khatatbeh, Lana Khatib, Salsabeel Qarariah, Yara Quzmar, Khalil Safadi, Roqaya Salameh, Mohammad Hassan, Shifaa Herzallah, Loai Massad, Ahmed Nazzal, Ranin Nazzal, Dennis Escobar, Gustavo Miguel Machain V, Agustin Rodriguez Gonzalez, Jorge Emerson Chachaima Mar, Nathaly Olga Chinchihualpa Paredes, Vicente Cuba, Walter Lopez, Maria Milagros Niquen Jimenez, Nestor Alberto Sanchez Bartra, Olenka Sapallanay Ojeda, Diego Sequeiros, Andrea Toscano Pacheco, María Vergara, Sol Abarca, Rodrigo Alcorta, Giuliano Borda-Luque, Ivan Edward Eusebio Zegarra, Claudia Luján López, Mirella Marrufo, Cinthya Mogrovejo, Andrea Nomura, Yamile Rodríguez Angeles, Maitza Rosario Vidal Meza, Gabriela Zavala, José Neiser Castillo Arrascue, Jomara Caroline Hidrogo Cabrera, José Julio Mariano Larrea vera, Miguel Osorio, Edgar Alcides Ylatoma Díaz, Mark Anthony Fontanilla, Joseph Roy Fuentes, Anna Leah Salazar, Genieve Dominguez, Marc Paul Lopez, Shiela Macalindong, Mark Augustine Onglao, Arjel Ramirez, Marie Dione Sacdalan, Mayou Martin Tampo, Gemma Leonora Uy, Jeremiah Mangahas, Kenneth Yabut, Joannes Paul Cañete, Bernalynn Eris Cansana, Ernes John Castro, Maria Kaiserin Lipana, Manuel Francisco Roxas, Vlu Jean Zara, Maciej Chroł, Paula Franczak, Michał Orłowski, Piotr Budzyński, Andrzej Budzyński, Pawel Bury, Agata Czerwińska, Jadwiga Dworak, Jacek Dziedzic, Michał Kisielewski, Jan Kulawik, Anna Lasek, Piotr Małczak, Marcin Migaczewski, Michał Pędziwiatr, Magdalena Pisarska, Dorota Radkowiak, Mateusz Rubinkiewicz, Anna Rzepa, Tomasz Skoczylas, Maciej Stanek, Katarzyna Truszkiewicz, Mateusz Wierdak, Marek Winiarski, Piotr Zarzycki, Anna Zub-Pokrowiecka, Piotr Kowalewski, Rafał Roszkowski, Maciej Walędziak, Miguel Tomé, Sara Patrocinio, Ines Guerreiro, Filipe Almeida, Xavier de Sousa, Nuno Monteiro, Maria Teresa Costa Santos, Daniela de Oliveira, Marta Lopes Serra, Daniela Morgado, Christian Neves, Ana Carolina Oliveira, Alice Pimentel, Sofia Silva, Márcia Carvalho, Lúcia Carvalho, Joana Magalhães, Leonor Matos, Tânia Monteiro, Carlota Ramos, Vanessa Santos, José Barbosa, Jose Costa-Maia, Vítor Devezas, Ana Fareleira, Cristina Fernandes, Diana Gonçalves, Henrique Mora, Marina Morais, Fabiana Silva de Sousa, Sara Catarino Santos, Ana Logrado, André Tojal, Edgar Amorim, Miguel F. Cunha, Ana Fazenda, João Pedro Melo Neves, Inês Isabel Sampaio da Nóvoa Gomes Miguel, Diogo Veiga, José Azevedo, Hugo Cardoso Louro, Mariana Leite, Maria Bairos Menezes, Bárbara Gama, Diana Brito, Marta Cristina Cruz Martins, André Graça e Magalhães, Ana Catarina Longras, Rita Lourenço, Diana Matos, Luis Castro, Filipa Policarpo, Joana Romano, Cristina Monteiro, Diogo Pinto, Marina Duarte, Sónia Fortuna Martins, Mariline Oliveira, Diogo Galvão, Lisandra Martins, Anaisa Silva, Viorel Taranu, Bárbara Vieira, Jessica Neves, Simone Oliveira, Hugo Ribeiro, Margarida Cinza, Rosa Felix, Arnaldo Machado, Joana Oliveira, Joana Patrício, Rita Pedroso de Lima, Mário Pereira, Miguel Rocha Melo, Cristina Velez, Alberto Abreu da Silva, Mariana Claro, Daniel Costa Santos, Andreia Ferreira, Hugo Capote, Daniela Rosado, Filipa Taré, Oriana Nogueira, Miguel Ângelo, José Miguel Baiao, Andreia Guimarães, João Marques, Miguel Nico Albano, Marta Silva, Ana Valente da Costa, Teresa Vieira Caroço, Sara Almeida Braga, Ines Capunge, Marta Fragoso, João Guimarães, Bruno Pinto, João Ribeiro, Miguel Angel, Guilherme Fialho, Monica Guerrero, Filipa Campos Costa, Diogo Cardoso, Vasco Cardoso, Magda Alves, Inês Estalagem, Tiago Louro, Cláudia Marques, Rita Martelo, Miguel Morgado, Rita Canotilho, Ana Margarida Correia, Pedro Martins, Mariana Peyroteo, João Gomes, Rita Monteiro, Manuela Romano, Daniela Macedo Alves, Rita Peixoto, Catarina Quintela, Maria João Jervis, Débora Melo, André Pacheco, Valter Paixão, Vera Pedro, Joana Pimenta, João Pimenta de Castro, Ana Rocha, Mircea Beuran, Cezar Ciubotaru, Bogdan Diaconescu, Sorin Hostiuc, Ionut Negoi, Bogdan Stoica, Evgeny Anokhin, Georgy Kuznetsov, Giorgi Oganezov, Fedor Paramzin, Ekaterina Romanova, Valeryan Rutkovskii, Vasilii Rutkovskii, Mikhail Shushval, Mikhail Zabiyaka, Khasan Dzhumabaev, Valerii Ivanov, Zaman Mamedli, Sergey Achkasov, Artem Balkarov, Elnur Nabiev, Marat Nagudov, Evgeny Rybakov, Karina Saifutdinova, Oleg Sushkov, Lule Joseph, Isaac Ndayishimiye, Ntirenganya Faustin, Alphonse Zeta Mutabazi, Jean Paul Mvukiyehe, Vizir J.P Nsengimana, Carine Uwakunda, Mohammad Monir Abbas, Nouf Akeel, Murad Aljiffry, Kholoud Awaji, Ali Farsi, Ghader Jamjoum, Ahmad Khoja, Ashraf Maghrabi, Nadim Malibary, Mohammed Nassif, Abdulaziz Saleem, Abdullah Sultan, Wail Tashkandi, Hanaa Tashkandi, Nora Trabulsi, Mouhamadou Bachir Ba, Adja Coumba Diallo, Abdourahmane Ndong, Vladica Cuk, Uroš Janković, Sharon Zhiling Koh, Frederick Koh, Kuok Chung Lee, Kai Yin Lee, Sean Lee, Wei Qi Leong, Su Ann Lui, Prajwala Prakash, Jan Grosek, Gregor Norcic, Ales Tomazic, Nicolas Fitchat, Robert Jaich, Devorah Wineberg, Modise Zacharia Koto, Daniella Baiocchi, Damian Clarke, Christina Johanna Steenkamp, Sharon Bannister, Adam Boutall, Galya Chinnery, Anna Coccia, Angela Dell, Parveen Karjiker, Christo Kloppers, Nicholas Loxton, Tumi Mabogoane, Francois Malherbe, Eugenio Panieri, Shreya Rayamajhi, Tirsa van Wyngaard, Claire Warden, T E Madiba, Nivashen Pillay, Savannah Brooks, Charlise Kruger, Lisa Hannah Van Der Merwe, Ferhana Gool, Maahir Kariem, Heather Bougard, Nazmie Kariem, Fazlin Noor, Reantha Pillay, Leandi Steynfaardt, Lucía González González, José Miguel Marín Santos, Paula Martín-Borregón, Javier Martínez Caballero, Cristina Nevado García, Pastora Rodriguez Fraga, Gonzalo De Castro Parga, Maria Pilar Fernández Veiga, Lucía Garrido López, Hugo Infante Pino, Irene Lages Cal, Marta López Otero, Manuel Nogueira Sixto, Marta Paniagua García Señorans, Laura Rodríguez Fernández, Alejandro Ruano Poblador, Erika Rufo Crespo, Raquel Sanchez-Santos, Vincenzo Vigorita, Ester Alonso Batanero, Dorisme Asnel, Isabel Cifrian Canales, Elisa Contreras Saiz, Irene De Santiago Alvarez, Tamara Díaz Vico, Sebastian Fernandez Arias, Daniel Fernández Martínez, Carmen García Bernardo, Luis Joaquín García Flórez, Carmen Garcia Gutierrez, Manuel García Munar, Carlos Alberto Márquez Zorrilla Molina, Marta Merayo, José Luis Michi Campos, Maria Moreno Gijon, Jorge L. Otero-Diez, Jose Luis Rodicio Miravalles, Lorena Solar-Garcia, Aida Suárez Sánchez, Nuria Truan, Cristina Alejandre Villalobos, Yurena Caballero Díaz, Marta Jimenez, Dacil Montesdeoca, Antonio Navarro-Sánchez, Victor Vega, Juan Beltrán de Heredia, Zahira Gómez, Carlos Jezieniecki, Ana Patricia Legido Morán, Mario Montes-Manrique, Mario Rodriguez-Lopez, María Ruiz Soriano, Jeancarlos Trujillo Díaz, Andrea Vazquez Fernandez, Nuria Argudo, Miguel Pera, Laia Torrent Jansà, Melody García Domínguez, Ignacio Goded, Marta Roldón Golet, Issa Talal El-Abur, Alejandra Utrilla Fornals, Vanesa Zambrana Campos, Maria Del Mar Aguilar Martinez, Marina Bosch, Luis García-Catalá, Luis Sánchez-Guillén, Eva Artigau, Nuria Gomez Romeu, David Julià Bergkvist, Beatriz Espina Perez, Olga Morató, Carles Olona, Beatriz Diéguez, Alexander Forero-Torres, Manuel Losada, Segundo Gomez-Abril, Paula Gonzálvez, Rosario Martinez, Sergio Navarro Martínez, Carmen Payá-Llorente, Álvaro Pérez Rubio, Sandra Santarrufina Martinez, Juan Carlos Sebastián Tomás, Ramon Trullenque Juan, Alberto Gegúndez Simón, Paloma Maté, Maria Isabel Prieto-Nieto, Ines Rubio-Perez, Aitor Urbieta, Marina Vicario Bravo, David Abelló, Matteo Frasson, Alvaro Garcia-Granero, Alfredo Abad Gurumeta, Ane Abad-Motos, Elena Lucena-de Pablo, Beatriz Nozal, Javier Ripollés-Melchor, Rut Salvachúa, Esther Ferrero, Luis Garcia-Sancho Tellez, Antonio L. Picardo, Jose Alberto Rojo López, Laura Patricia Zorrilla Matilla, Carmen Cagigas Fernandez, Sonia Castanedo Bezanilla, José Estevez Tesouro, Maria Jose Fernandez-Diaz, Juan García Cardo, Marcos Gomez Ruiz, Erik Gonzalez-Tolaretxipi, Jaime Jimeno Fraile, Cristobal Poch, Montserrat Rodriguez-Aguirre, Noemí Troche Pesqueira, Maria Soledad Trugeda-Carrera, Javier de la Torre, Ruth Blanco-Colino, Eloy Espin-Basany, Martin Espinosa-Bravo, Clara Morales Comas, Eduardo Reyes Afonso, Joaquín Rivero Déniz, Christian Siso Raber, Mireia Verdaguer Tremolosa, Pramodh Chandrasinghe, Sumudu Kumarage, Nimeshi Wijekoon Arachchilage, Ahmed Abdalla Ahmed Elkamel, Mohammed A. Adam, Nina Blomme, Anders Thorell, Fredrik Wogensen, Andreas Älgå, Dhirar Ansarei, Fuat Celebioglu, Göran Heinius, Linda Nigard, Emil Pieniowski, Sandra Ahlqvist, Ida Björklund, Andreas Frånberg, Martina Håkansson, Karin Adamo, Oskar Franklin, Malin Sund, Rebecca Wiberg, Yvette Andersson, Abbas Chabok, Maziar Nikberg, Alexander Kugelberg, Claudia Canonica, Dimitrios Christoforidis, Fabrizio Fasolini, Paolo Gaffuri, Mauro Giuliani, Francesco Meani, Sotirios Georgios Popeskou, Silvia Pozza, Wiebke Wandschneider, Lorenz Peterer, Lukas Werner Widmer, Bernd Zimmermann, Panagiotis Bakoleas, Iris Chanousi, Lydia Charalampidou, Lukasz Filip Grochola, Franziska Heid, Sotirios Ntaoulas, Michail Outos, Georgios Peros, Hanna Podolska-Skoczek, Katharina Beate Reinisch, Christian Zielasek, Nicolas Demartines, Jérôme Gilgien, Amaniel Kefleyesus, Pénélope St-Amour, Arnaud Toussaint, Maryam Alhimyar, Bayan Alsaid, Amr Alyafi, Ahmad Alkhaledi, Basel Kouz, Ahmad Omarain, Yusra Al-Sabbagh, Haya Alkhatib, Samer Sara, Ahmad Alhaj, Aghyad Danial, Lama Kadoura, Sarah Maa Albared, Yamen Monawar, Louei Nahas, Barook Abd, Ahmad Saad, Habib Wakkaf, Hatem Bouzaiene, Montassar Ghalleb, Elif Akaydin, Ata Cem Akbaba, Onur Atakul, Ege Baltaci, Sevval Besli, Gökçen Burgu, Ulukan Cenal, Cansu de Muijnck, Hasan Can Demirkaya, Alper Dogruoz, Zeynep Ipek Gezer, Yasemin Gündoğdu, Merve Kara, Hasan Kürşad Korkmaz, Gökalp Kağan Kurtoğlu, Volkan Ozben, Berk Baris Ozmen, Ahmet Murat Pektaş, Eda Kübra Sel, Nilüfer Yenidünya, Fuat Baris Bengur, Berke Mustafa Oral, Tahir Koray Yozgatli, Seymur Abdullayev, Mehmet Emin Gunes, Nuri Alper Sahbaz, Tuba Banaz, Kübra Kargici, Omer Faruk Kuyumcu, Erkan Yanikoğlu, Merve Yeşilsancak, Duygu Yilmaz, Melik Kagan Aktas, Ahmet Rencuzogullari, Arda Isik, Sezai Leventoğlu, Ali Yalçinkaya, Osman Yüksel, Mustafa U Kalayci, Yasin Kara, Inanc Samil Sarici, Alp Akin, Gökçe nur Alemdağ, Ekin Arslan, Bahadir Emre Baki, Muhammed Selim Bodur, Adnan Calik, Bahar Candas Altinbas, Irem Cihanyurdu, Oğuz Erkul, Burak Gül, Ali Guner, Beyza Köse, Anil Semiz, Şule Sevim, Serkan Tayar, Kadir Tomas, Ozan yavuz Tüfek, Serdar Türkyilmaz, Mehmet Uluşahin, Arif Usta, Reyyan Yildirim, Sertaç Ata Güler, Ozan Can Tatar, Ecenur Varol, Busenur Kirimtay, Muhammed Uysal, Alp Yildiz, Emin Kose, Ahmet Burak Ciftci, Elif Çolak, Huseyin Eraslan, Gultekin Ozan Kucuk, Kürşat Yemez, Herman Lule, Mumbere Bienfait, Emmanuel Bua, Noella Okalany, Maksym Basarab, Oleksii Bielosludtsev, Kateryna Kolhanova, Kateryna Perepelytsia, Kateryna Romanukha, Dmytro Savenkov, Stanislav Siryi, Maksym Tereshchenko, Nezamai Viacheslav, Anton Volovetskyi, Andrey Kebkalo, Yegor Tryliskyy, Volodimir Tyselskiy, Eilidh Bruce, Bing Lun Chow, Emma Iddles, Sarah McGuckin, Nicola Newall, George Ramsay, Parivrudh Sharma, Caitlin Stewart, Jeremy Wong, Abdul Badran, Michael Bath, Fanny Belais, Eman Butt, Kaustuv Joshi, Milan Kapur, Mike Shaw, Adam Townson, Christopher Yee Khang Williams, Timothy Gray, Robert Greig, Mansoor Husain, Elspeth Murray, Ahmed Mustafa, Ashar Asif, Arya Gokul, Max Shah, Mabel Temisanren Akitikori, Alexandros Charalabopoulos, Sophie Davidson, Sinead McNally, Shamil Rupani, Fatema Juma, Sarah Catherine Mills, Laura Muirhead, Kate Sellars, Una Walsh, Oliver Warren, Alice Chambers, Richard Hunt, Stephen Boyce, Hannah Cornwall, Isabel Tol, Eleftherios Orestis Argyriou, Nicola Eardley, Meical Povey, Joanna M S Aithie, Ahmer Irfan, Mari-Claire McGuigan, Robert Starr, Craig Russell Warren, Jess Archibald, Georgia Kirby, Ivan Kisyov, Chun Kheng Khoo, Rachel Lee, Dana Photiou, Rowan Davis, Uday Prasad, P Zichu Yang, Jonathan Bird, Edmund Leung, Virginia Summerour, Chelise Currow, Jianshen Kiam, Gerald Jack Soon Tan, Anitha Muthusami, Ibifunke Pegba-Otemolu, Tomas Urbonas, Joseph Nunoo-Mensah, Edgaras Smolskas, Alex Boddy, Gianpiero Gravante, David Hunter, David Andrew, Amanda Koh, Amari Thompson, Lawrence Adams, Hollie A Clements, Kasun De Silva, Ogbonnia Ekpete, Seraj Haque, Scott Henderson, Bilal Ibrahim, Thummini Jayasinghe, Jennifer Livie, Keir Mailley, Gopikrishnan Nair, Daniel Tan, Caitlin Baggaley, Aleksander Dawidziuk, Bartosz Szyszka, Charlotte Barter, Nirav Gandhi, Karen Hassell, Samantha Hitchin, Jennett Kelsall, Eva Nagy, Ashrafun Nessa, Lisa Whisker, Fady Yanni, Mahmoud Ali, Deeksha Arora, Sunanda Hediwattege, Navam Kumarasinghe, Munir Rathore, Athula Tennakoon, Syed Mustafa Ali Ahmad, Oreoluwa Bajomo, Fahema Nadira, Valerio Celentano, Ewen Griffiths, Rama Santhosh Karri, Jason Kei Chak Mak, Michelle Pipe, Muhammad Iqbal Bhatti, Mohamed Rabie, Connor Boyle, David Hamilton, Aishath Mihuna, James Chean Khun Ng, Gary Nicholson, Agata Oliwa, Robert Pearson, Anna Rose, Shun Qi Yong, Catherine Boereboom, Michael Hanna, Catherine Walter, Thomas Samuel Greensmith, Rachel Mitchell, Eimear Monaghan, James Crawford, Susan Moug, James Blackwell, Hannah Boyd-Carson, Philip Herrod, Omar Al-Allaf, Miriam Beattie, Cameron Bullock, Shivang Burman, Gemma Clark, Nicolas Flamey, Oliver Flannery, Alexander Harding, Ben Kodiatt, Samuel Lawday, Shivani Mahapatra, Navin Mukundu Nagesh, Michael Ng, Dupinderjit Rye, Andrel Yoong, Laura Clark, Chris Deans, Monisha Edirisooriya, Emma Victoria Carrington, Tsz Lun Ernest Wong, Baasil Yusuf, Carla Chamberlain, Kathryn Duke, Elizabeth Kmiotek, Azel Botes, Natalie Condie, Timothy Schrire, Reena Shah, Iolo Thomas-Jones, Charlotte Yates, Natasha Anthony, Edward Matthews, Kapil Sahnan, James Tankel, Sally Tucker, Jasmine Winter Beatty, Paul Ziprin, William Duggan, Anastasia Kantartzi, Shruthi Sridhar, Rachel Alys Khaw, Prakhar Srivastava, Charlotte Underwood, Homero Alves do Canto Brum, Sharat Chopra, Laura Davis, Rebecca Hughes, Joshua Tulley, Justin Alberts, Thomas Athisayaraj, Mojolaoluwa Olugbemi, Kasim Ahmad, Claudia Chan, Gavin Chapman, Hannah Fleming, Benjamin Fox, Julia Grewar, Kate Hulse, Duncan Rutherford, Mackay Sinead, Scott Smith, Doug Speake, Peter G Vaughan-Shaw, Natasha Christodoulides, Simrit Kudhail, Matthew Welch, Syed Muhibullah Husaini, Simon Lambracos, Chikamuche Anyanwu, Rishi Suresh, Jimmy Scott Thomas, Elizabeth Gleeson, Rebecca Platoff, Areeba Saif, Zachary Enumah, Eric Etchill, Alodia Gabre-Kidan, Mitchell Bernstein, Francesco Maria Carrano, Joseph Connors, Patricio Lynn, Marcovalerio Melis, Elliot Newman, Deshka S Foster, Kenneth Perrone, Ashley Titan, Sarwat Ahmad, Andrea Chao M.D. Bafford, Marco Dal Molin, Nader Hanna, Syed Nabeel Zafar, Mark Hemmila, Lena Napolitano, Jane J Wong, Julia Chandler, Lauren Wood, Sherry Wren, Taylor Ottesen, Lucia You, Kristin Yu, María del pilar Arciénega Yañez, Martin Ferreira Fernandes, Daniel González, Santiago Cubas, María Catalina González, Vanessa Zubiaurre, Rodrigo Demolin, Nicolas Giroff, Pablo Sciuto, Maite Campos, Gabriela Rodríguez Cantera, Garg Deepika, Elliot Simuchimba, Anadi Bulaya, Chali Chibuye, Bright Chirengendure, Mary-Rose Kabale, Kizito Kabongo, James Munthali, Oliver Mweso, Francis Pikiti, James Otieno, Log Tung Lai, Brighid Blackman, Sophie Richards, Suren Subramaniam, Rafid Karim, Nathan Kok, Yanni Dion Lee, Shabina Ali, Aanjaneya Sinha, Robert Corrigan, Nicole Barnes, Florence Wong, Grace Dennis, Julia Jedamzik, Emil Phillips, Wivine Piette, Marie Van hentenryck, Houenoukpo Koco, Souliath Lawani, Mamo Woldu Kassa, Tainá Santos Bezerra, Petar Gribnev, Dobromir Dimitrov, Panche Krastev, Sovannarith Oum, Divine Tim Bonghaseh, Maryam Al Farsi, Nourah Alsharqawi, Veronica Acevedo, Andrea Carolina Castillo Barbosa, Felipe Giron, Jimmy Paul Leon Rodriguez, Darko Kučan, Damir Rosko, Neven Barsic, Domagoj Župan, Amgad Hegazi, Vendula Trunčíková, Vladimir Fryba, Mostafa Mohamed, Ahmed Sultan, Ahmed Nagi, Abdallah Rashad Temerik, Mohamed Elemam Elshawy, Moustafa Ibrahim Mahmoud, Shrouk Omar, Mohamed Anwar, Tarek Rageh, Aya Elmokadem, Khaled Gaballa, Sandra Teppo, Antti Turunen, Pasi Pengermä, Quentin Ballouhey, Damien Bergeat, Ariane Weyl, Elisabeth Hain, Adam Gyedu, Edwin Yenli, Dorcas Osei-Poku, Vaia-Aliki Rompou, Athanasios Zoikas, Apostolos Gaitanidis, Georgios Koukis, Konstantinos Perivoliotis, Panagiotis Tavlas, Konstantinos Galanos-Demiris, George Zografos, Ioannis Karavokyros, Georgia Xanthopoulou, Eirini Iordanidou, Fernanda Ayau, Allan Garcia, Pekli Damján, Deepender Wason, Ashika B L, Ervandy Rangganata, Prerna Kamath, Donal B O'Connor, Margherita Pinto, Fabrizio Perrone, Francesca Paola Tropeano, Francesca Troilo, Daniela Bossi, Dario Scala, Lucrezia Pulitanò, Marcella Carella, Andrea Pietrabissa, Alice Gori, Giorgio Giraudo, Veronica De Simone, Alfio Alessandro Russo, Bartolomeo Braccio, Raed Al-Taher, Sarah Athamneh, Andrea Parker, Adnan Sawiee, Amina Kattia, Malik Salem, Osama Tababa, Zuhour Shaeeb, Vilius Syminas, Jonas Jurgaitis, Gytè Damulevičienè, Saulius Svagzdys, Narindra Njarasoa Mihaja Razafimanjato, Ling Chieng Loo, Ing Ching Tiong, Wan Farahiyah Wan Muhmad, Harinthiran Vijeyan, Teoh Li Ying, Gabriella Grech, Rodrigo Arrangoiz, Vania Brickelia Jimenez Ley, Daniel Arizpe, Elizabeth Lagunes Lara, Elizabeth Victoria Castro López, Jose Eaazim, Marije Gordinou de Gouberville, Vivian Bastiaenen, Simone Rottier, Fouad Nahab, Maria Yeonhee Ji, Mohammed Seyoji, Callistus Nwachukwu, Okechukwu Emeghara, Sayyid Egbunu Muhammed, Ayodeji Idowu, Olamiposi Sowemimo, Olakayode Ogundoyin, Oluwatosin Akande, Alexander Lott, Maliha Nadeem, Ahsan Ali Laghari, Asif Loya, Hassan Mushtaq, Muhammad Tariq Abdullah, Baseel Abuhilal, Mohammad Atawneh, Hamdan Hamdan, Belal Alhabil, Abedelrahman Srour, Ibrahim Mousa, Luis Da Silva Medina, Katarzyna Bartosiak, Pedro Ferreira, Vítor Francisco, Ricardo Lemos, Luísa Frutuoso, Sara Fernandes, Telma Fonseca, Jorge Pereira, Juan Rachadell, Ana Torre, Filipe Madeira Martins, Ana Cristina Carvalho, Joana Rodrigues Ferreira, Bruno Ribeiro da Silva, Helena Devesa, Ana Vieira, Inês Mónica, Margarida Amaro, Diogo Sousa, Marta Reia, João Louro, Ana Martins, Joaquina Dominguez, Inês Santos, Nuno Miguel Freitas Oliveira, José Carlos Pereira, Pedro Silva-Vaz, Ligia Freire, Ricardo Escrevente, Valentina Madalina Negoita, Dmitry Shakhmatov, Yves Nezerwa, Radosav Radulovic, Gareth Obery, Francois Viljoen, Tome Mendes, Antonio Suarez, Enrique Moncada, Maria Fernandez-Hevia, Carolina Curtis Martínez, Julia Maria Gil Garcia, Mariana González Zunzarren, Tarig Idris, Karolina Eklöv, Oskar Grahn, Leila Amin, Malin Blomqvist, Costanza Ajani, Rebecca Kraus, Nico Seeger, Melissa Willemin, Fadi Rayya, Mohammad Ayash, Raneem Msouti, Israa Kannas, Eias Abazid, Asil Esper, Skander Slim, Akil Serdar Kavcar, Erman Aytac, Ahmet Cem Dural, Ayse Ilker, Ismail Cem Eray, Eray Kurnaz, Saygin Altiner, Mustafa Deniz Tepe, Can Şahin, Evrim Savli, Aryon Innocent, Lilian Babirye, Andrii Diachenko, Vladislav Hordoskiy, Heather Curry, Charlene Yat Che Chau, Harry Robertson, Arin Mahmoud, Hannah Lennon, Lynette Loi, Emily Kirkham, Cameron McCann, Daniel Watts, Binay Gurung, Michael Wilson, Thomas Tribedi, Eleonora Garofalo, Baryab Zahra, Scott MacDonald, Ian Daniels, Nathan Ng, Shivun Khosla, James Olivier, Sum Yu Pansy Yue, Gayathri Suresh, Jack Wellington, Emmanuel Lorejo, Mafdi Mossaad, Madison Crutcher, Marjan Alimi, Ioana Baiu, Hossam Abdou, Alison Conway, Connor Peck, Mauro Andres Perdomo Perez, Stanley Zulu, Mildred Nakazwe, Sule Burger, Justine Davies, Rachel Donaldson, Chikwendu Ede, O James Garden, Chiapo Lesetedi, Charles Mabedi, Laura Magill, Felix Makinde Alakaloko, Alex Makupe, Mark Monahan, Soloman Mulira, Elmi Muller, Jospeh Musowoyo, Jean Léon Olory-Togbe, Tracey Roberts, Martin Smith, Viki Tayler, John Windsor, Raul Yepez, Sudha Sundar, Emmy Runigamugabo, Azmina Verjee, José Chen, Leonid Daya, Nouhaila El Aroussi, Valeria Farina, Tchianze Gnintedeme Olivier, Mauricio Gonzales Nacarino, Aamr Hammani, Sarah Honjo, Rebecca Jacobs, Hitomi Kimura, Mugisha Nkoronko, Jasson Javier Oscullo Yepez, Wei Pin Hung, Ankit Raj, Alina Romani Pozo, Muna Rommaneh, Samuel Chimbioputo Sassamela Fabiano, Camila Milagros Shiroma Gago, Abhishekh Srinivas, Chia-Yen Sung, Aswan Tai, Yener Cristyell Valle Aranda, Sara Venturini, and Jean Wilguens Lartigue

Subjects
Public aspects of medicine, RA1-1270
Abstract

Summary: Background: Early death after cancer surgery is higher in low-income and middle-income countries (LMICs) compared with in high-income countries, yet the impact of facility characteristics on early postoperative outcomes is unknown. The aim of this study was to examine the association between hospital infrastructure, resource availability, and processes on early outcomes after cancer surgery worldwide. Methods: A multimethods analysis was performed as part of the GlobalSurg 3 study—a multicentre, international, prospective cohort study of patients who had surgery for breast, colorectal, or gastric cancer. The primary outcomes were 30-day mortality and 30-day major complication rates. Potentially beneficial hospital facilities were identified by variable selection to select those associated with 30-day mortality. Adjusted outcomes were determined using generalised estimating equations to account for patient characteristics and country-income group, with population stratification by hospital. Findings: Between April 1, 2018, and April 23, 2019, facility-level data were collected for 9685 patients across 238 hospitals in 66 countries (91 hospitals in 20 high-income countries; 57 hospitals in 19 upper-middle-income countries; and 90 hospitals in 27 low-income to lower-middle-income countries). The availability of five hospital facilities was inversely associated with mortality: ultrasound, CT scanner, critical care unit, opioid analgesia, and oncologist. After adjustment for case-mix and country income group, hospitals with three or fewer of these facilities (62 hospitals, 1294 patients) had higher mortality compared with those with four or five (adjusted odds ratio [OR] 3·85 [95% CI 2·58–5·75]; p

Published
2022
Full Text
View/download PDF

6. Adversarial Training in Low-Label Regimes with Margin-Based Interpolation

Author

Ye, Tian, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning, Computer Science - Cryptography and Security, Computer Science - Computer Vision and Pattern Recognition
Abstract

Adversarial training has emerged as an effective approach to train robust neural network models that are resistant to adversarial attacks, even in low-label regimes where labeled data is scarce. In this paper, we introduce a novel semi-supervised adversarial training approach that enhances both robustness and natural accuracy by generating effective adversarial examples. Our method begins by applying linear interpolation between clean and adversarial examples to create interpolated adversarial examples that cross decision boundaries by a controlled margin. This sample-aware strategy tailors adversarial examples to the characteristics of each data point, enabling the model to learn from the most informative perturbations. Additionally, we propose a global epsilon scheduling strategy that progressively adjusts the upper bound of perturbation strengths during training. The combination of these strategies allows the model to develop increasingly complex decision boundaries with better robustness and natural accuracy. Empirical evaluations show that our approach effectively enhances performance against various adversarial attacks, such as PGD and AutoAttack.

Published
2024

7. Conformal Prediction for Federated Graph Neural Networks with Missing Neighbor Information

Author

Akgül, Ömer Faruk, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning
Abstract

Graphs play a crucial role in data mining and machine learning, representing real-world objects and interactions. As graph datasets grow, managing large, decentralized subgraphs becomes essential, particularly within federated learning frameworks. These frameworks face significant challenges, including missing neighbor information, which can compromise model reliability in safety-critical settings. Deployment of federated learning models trained in such settings necessitates quantifying the uncertainty of the models. This study extends the applicability of Conformal Prediction (CP), a well-established method for uncertainty quantification, to federated graph learning. We specifically tackle the missing links issue in distributed subgraphs to minimize its adverse effects on CP set sizes. We discuss data dependencies across the distributed subgraphs and establish conditions for CP validity and precise test-time coverage. We introduce a Variational Autoencoder-based approach for reconstructing missing neighbors to mitigate the negative impact of missing data. Empirical evaluations on real-world datasets demonstrate the efficacy of our approach, yielding smaller prediction sets while ensuring coverage guarantees.

Published
2024

8. Simulating the blood transfusion system in Kenya: Modelling methods and exploratory analyses

Author

Tian, Yiqi, Zeng, Bo, MacLeod, Jana, Murithi, Gatwiri, Makanga, Cindy M., Barmasai, Hillary, Barnes, Linda, Bidanda, Rahul S., Epuu, Tonny Ejilkon, Kaburu, Robert Kamu, Chelagat, Tecla, Madan, Jason, Makin, Jennifer, Munoz-Valencia, Alejandro, Njoki, Carolyne, Ochieng, Kevin, Olayo, Bernard, Paiz, Jose, Rudd, Kristina E., Yazer, Mark, Puyana, Juan Carlos, Bidanda, Bopaya, Rajgopal, Jayant, and Kumar, Pratap

Subjects
Physics - Medical Physics, Electrical Engineering and Systems Science - Systems and Control
Abstract

The process of collecting blood from donors and making it available for transfusion requires a complex series of operations involving multiple actors and resources at each step. Ensuring hospitals receive adequate and safe blood for transfusion is a common challenge across low- and middle-income countries, but is rarely addressed from a system level. This paper presents the first use of discrete event simulation to study the blood system in Kenya and to explore the effect of variations and perturbations at different steps of the system on meeting patient blood demand. A process map of the Kenyan blood system was developed to capture critical steps from blood donation to transfusion using interviews with blood bank, hospital, and laboratory personnel at four public hospitals across three counties in Kenya. The blood system was simulated starting with blood collection, a blood bank where blood is tested and stored before it is issued, a major hospital attached to the blood bank, and several smaller hospitals served by the same blood bank. Values for supply-side parameters were based mainly on expert opinion; demand-side parameters were based on data from blood requisitions made in hospital wards, and dispatch of blood from the hospital laboratory. Illustrative examples demonstrate how the model can be used to explore the impacts of changes in blood collection (e.g., prioritising different donor types), blood demand (e.g., differing clinical case mix), and blood distribution (e.g., restocking strategies) on meeting demand at patient level. The model can reveal potential process impediments in the blood system and aid in choosing strategies for improving blood collection, distribution or use. Such a systems approach allows for interventions at different steps in the blood continuum to be tested on blood availability for different patients presenting at diverse hospitals across the country., Comment: 38 pages, 8 figures

Published
2024

9. Robust Quickest Correlation Change Detection in High-Dimensional Random Vectors

Author

Alghamdi, Assma, Banerjee, Taposh, and Rajgopal, Jayant

Subjects
Statistics - Methodology, Electrical Engineering and Systems Science - Signal Processing
Abstract

Detecting changes in high-dimensional vectors presents significant challenges, especially when the post-change distribution is unknown and time-varying. This paper introduces a novel robust algorithm for correlation change detection in high-dimensional data. The approach utilizes the summary statistic of the maximum magnitude correlation coefficient to detect the change. This summary statistic captures the level of correlation present in the data but also has an asymptotic density. The robust test is designed using the asymptotic density. The proposed approach is robust because it can help detect a change in correlation level from some known level to unknown, time-varying levels. The proposed test is also computationally efficient and valid for a broad class of data distributions. The effectiveness of the proposed algorithm is demonstrated on simulated data.

Published
2024

10. CROSS-GAiT: Cross-Attention-Based Multimodal Representation Fusion for Parametric Gait Adaptation in Complex Terrains

Author

Seneviratne, Gershom, Weerakoon, Kasun, Elnoor, Mohamed, Rajgopal, Vignesh, Varatharajan, Harshavarthan, Jaffar, Mohamed Khalid M, Pusey, Jason, and Manocha, Dinesh

Subjects
Computer Science - Robotics
Abstract

We present CROSS-GAiT, a novel algorithm for quadruped robots that uses Cross Attention to fuse terrain representations derived from visual and time-series inputs, including linear accelerations, angular velocities, and joint efforts. These fused representations are used to adjust the robot's step height and hip splay, enabling adaptive gaits that respond dynamically to varying terrain conditions. We generate these terrain representations by processing visual inputs through a masked Vision Transformer (ViT) encoder and time-series data through a dilated causal convolutional encoder. The cross-attention mechanism then selects and integrates the most relevant features from each modality, combining terrain characteristics with robot dynamics for better-informed gait adjustments. CROSS-GAiT uses the combined representation to dynamically adjust gait parameters in response to varying and unpredictable terrains. We train CROSS-GAiT on data from diverse terrains, including asphalt, concrete, brick pavements, grass, dense vegetation, pebbles, gravel, and sand. Our algorithm generalizes well and adapts to unseen environmental conditions, enhancing real-time navigation performance. CROSS-GAiT was implemented on a Ghost Robotics Vision 60 robot and extensively tested in complex terrains with high vegetation density, uneven/unstable surfaces, sand banks, deformable substrates, etc. We observe at least a 7.04% reduction in IMU energy density and a 27.3% reduction in total joint effort, which directly correlates with increased stability and reduced energy usage when compared to state-of-the-art methods. Furthermore, CROSS-GAiT demonstrates at least a 64.5% increase in success rate and a 4.91% reduction in time to reach the goal in four complex scenarios. Additionally, the learned representations perform 4.48% better than the state-of-the-art on a terrain classification task.

Published
2024

11. Learning Personalized Scoping for Graph Neural Networks under Heterophily

Author

Deng, Gangda, Zhou, Hongkuan, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning, Computer Science - Social and Information Networks
Abstract

Heterophilous graphs, where dissimilar nodes tend to connect, pose a challenge for graph neural networks (GNNs) as their superior performance typically comes from aggregating homophilous information. Increasing the GNN depth can expand the scope (i.e., receptive field), potentially finding homophily from the higher-order neighborhoods. However, uniformly expanding the scope results in subpar performance since real-world web graphs often exhibit homophily disparity between nodes. An ideal way is personalized scopes, allowing nodes to have varying scope sizes. Existing methods typically add node-adaptive weights for each hop. Although expressive, they inevitably suffer from severe overfitting. To address this issue, we formalize personalized scoping as a separate scope classification problem that overcomes GNN overfitting in node classification. Specifically, we predict the optimal GNN depth for each node. Our theoretical and empirical analysis suggests that accurately predicting the depth can significantly enhance generalization. We further propose Adaptive Scope (AS), a lightweight approach that only participates in GNN inference. AS encodes structural patterns and predicts the depth to select the best model for each node's prediction. Experimental results show that AS is highly flexible with various GNN architectures across a wide range of datasets while significantly improving accuracy.

Published
2024

12. Therapeutic Potential of Cannabidiol (CBD) for Skin Health and Disorders

Author

Baswan SM, Klosner AE, Glynn K, Rajgopal A, Malik K, Yim S, and Stern N

Subjects
cannabidiol (cbd), cannabinoids, endocannabinoids, endocannabinoid system, cb1, cb2, faah, aea., Dermatology, RL1-803
Abstract

Sudhir M Baswan,1,* Allison E Klosner,2,* Kelly Glynn,1 Arun Rajgopal,1 Kausar Malik,1 Sunghan Yim,1 Nathan Stern1 1Innovation and Science, Amway Corporation, Ada, MI, 49355, USA; 2Innovation and Science, Nutrilite Health Institute, Amway Corporation, Buena Park, CA, 90621, USA*These authors contributed equally to this workCorrespondence: Sudhir M Baswan Amway Corporation, 7575 Fulton St E, Ada, MI 49355, USATel +1-616-787-0216Email sudhir.baswan@amway.comAbstract: Though there is limited research confirming the purported topical benefits of cannabinoids, it is certain that cutaneous biology is modulated by the human endocannabinoid system (ECS). Receptors from the ECS have been identified in the skin and systemic abuse of synthetic cannabinoids, and their analogs, have also been associated with the manifestation of dermatological disorders, indicating the effects of the ECS on cutaneous biology. In particular, cannabidiol (CBD), a non-psychoactive compound from the cannabis plant, has garnered significant attention in recent years for its anecdotal therapeutic potential for various pathologies, including skin and cosmetic disorders. Though a body of preclinical evidence suggests topical application of CBD may be efficacious for some skin disorders, such as eczema, psoriasis, pruritis, and inflammatory conditions, confirmed clinical efficacy and elucidation of underlying molecular mechanisms have yet to be fully identified. This article provides an update on the advances in CBD research to date and the potential areas of future exploration.Keywords: cannabidiol, CBD, cannabinoids, endocannabinoids, endocannabinoid system, skin, CB1, CB2, FAAH, AEA

Published
2020

13. A pre-post trial to examine biological mechanisms of the effects of time-restricted eating on symptoms and quality of life in bipolar disorder.

Author

Johnson, Sheri, Murray, Greg, Manoogian, Emily, Mason, Liam, Allen, J, Berk, Michael, Panda, Satchidananda, Rajgopal, Nandini, Gibson, Jake, Bower, Carter, Berle, Eline, Joyner, Keanan, Villanueva, Robert, Michalak, Erin, and Kriegsfeld, Lance

Subjects
Bipolar disorder, Circadian rhythms, Depression, Diurnal, Intervention, Mania, Metabolic health, Quality of life, Sleep, Time-restricted eating, Humans, Bipolar Disorder, Quality of Life, Adult, Female, Male, Sleep Wake Disorders, Fasting
Abstract

BACKGROUND: The primary objective of this trial is to examine the mechanisms of time-restricted eating (TRE) as an adjunct to psychiatric care for people with bipolar disorder (BD) with sleep or circadian disruptions. This study builds on prior studies of circadian disruption in BD as well as growing evidence that TRE improves circadian functioning. METHODS: One-hundred fifty participants diagnosed with BD 1 or II will be recruited via advertising in the local community. Main inclusion criteria include: obtaining medical treatment for BD; current sleep or circadian problems; self-reported eating period of ≥ 12 h; no eating disorder or other health conditions that would hinder or limit the safety of following TRE; and not currently experiencing a mood episode, acute suicidality, psychosis, alcohol or substance use disorder. Participants will be asked to complete a baseline period in which daily food intake is logged online for two weeks. After baseline, participants will be asked to follow TRE for 8 weeks and to continue to complete daily food logging during this time. Symptom severity interviews will be conducted by phone or videoconference at baseline, mid-intervention (6 weeks post-baseline), end of intervention (10 weeks post-baseline), and 6 months post-baseline. Self-rated symptom severity and quality of life data will be gathered online at the same time points as symptom severity interviews, and at 16 weeks post-baseline (6 weeks after the TRE period ends). To assess potential mechanisms of change, we will examine the change in diurnal amplitude of clock gene expression as a primary mediator at 8 weeks compared to baseline. We will further test whether diurnal amplitude of clock gene expression is predictive above and beyond the role of two covariate potential mediators, glucose tolerance and inflammation at 8 weeks relative to baseline. To provide an index of whether TRE successfully decreases emotional lability, participants will be asked to complete 5 mood assessments per day for 7 days at baseline and at 10 weeks. These mood assessments will be optional. DISCUSSION: The planned research will provide novel and important information on whether TRE improves sleep/circadian rhythm problems, along with reductions in mood symptoms and improvements in quality of life, for individuals with BD. TRIAL REGISTRATION: ClinicalTrials.gov ID: NCT06555406.

Published
2024

14. Studying the Effects of Self-Attention on SAR Automatic Target Recognition

Author

Fein-Ashley, Jacob, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Attention mechanisms are critically important in the advancement of synthetic aperture radar (SAR) automatic target recognition (ATR) systems. Traditional SAR ATR models often struggle with the noisy nature of the SAR data, frequently learning from background noise rather than the most relevant image features. Attention mechanisms address this limitation by focusing on crucial image components, such as the shadows and small parts of a vehicle, which are crucial for accurate target classification. By dynamically prioritizing these significant features, attention-based models can efficiently characterize the entire image with a few pixels, thus enhancing recognition performance. This capability allows for the discrimination of targets from background clutter, leading to more practical and robust SAR ATR models. We show that attention modules increase top-1 accuracy, improve input robustness, and are qualitatively more explainable on the MSTAR dataset.

Published
2024

16. CAGI6 ID panel challenge: assessment of phenotype and variant predictions in 415 children with neurodevelopmental disorders (NDDs)

Author

Aspromonte, Maria Cristina, Del Conte, Alessio, Zhu, Shaowen, Tan, Wuwei, Shen, Yang, Zhang, Yexian, Li, Qi, Wang, Maggie Haitian, Babbi, Giulia, Bovo, Samuele, Martelli, Pier Luigi, Casadio, Rita, Althagafi, Azza, Toonsi, Sumyyah, Kulmanov, Maxat, Hoehndorf, Robert, Katsonis, Panagiotis, Williams, Amanda, Lichtarge, Olivier, Xian, Su, Surento, Wesley, Pejaver, Vikas, Mooney, Sean D., Sunderam, Uma, Srinivasan, Rajgopal, Murgia, Alessandra, Piovesan, Damiano, Tosatto, Silvio C. E., and Leonardi, Emanuela

Published
2025
Full Text
View/download PDF

19. Pancreaticojejunostomy for Pancreatico-enteric Anastomosis after Pancreaticoduodenectomy: one procedure with multiple techniques.

Author

Rajesh S Shinde, Rajgopal Acharya, Vikram A Chaudhari, Manish S Bhandare, and Shailesh V Shrikhande

Subjects
Pancreatico-jejunostomy, Pancreatico-enteric anastomosis, Pancreaticoduodenectomy, Surgery, RD1-811
Abstract

Pancreatico-enteric anastomoses after Pancreatico-duodenectomy (PD) has remained a true Achilles' heel in this complex and technically demanding operation. Multiple techniques have been described to achieve safe pancreatico-enteric drainage, however postoperative pancreatic fistula (POPF), still remains the major cause of morbidity. Pancreatico-jejunostomy (PJ) is commonly performed procedure for this purpose, however there are lot of variations in PJ techniques. Herein, we intend to review the commonly practised PJ techniques with emphasis on technical variations and reported outcomes. Literature review suggest that, there is no single, universally accepted or gold standard technique for this purpose. Pancreatic surgeons should be aware of technical aspects of different techniques, so that they can choose the best suited technique as per the clinical scenario.

Published
2020
Full Text
View/download PDF

20. Analysis of Secondary Surgeries after Immediate Breast Reconstruction for Cancer Compared with Risk Reduction

Author

Stacey J. Jones, MRCS(Ed), MBChB, MPharm, Philip Turton, FRCS(Ed), MD, Rajgopal Achuthan, FRCS(I), PhD, MBBS, Brian V. Hogan, FRCS, MD, MBBChBAO, Shireen N. Mckenzie, FRCS(Ed), MD, PGCert(ClinEd) FHEA, and Baek Kim, FRCS, MD, MA

Subjects
Surgery, RD1-811
Abstract

Background:. This study sets out to compare reconstructive practice between patients undergoing immediate breast reconstruction (IBR) for cancer and those who opted for risk reduction (RR), with an emphasis on examining patterns of secondary surgery. Methods:. Data collection was performed for patients undergoing mastectomy and IBR at a teaching hospital breast unit (2013–2016). Results:. In total, 299 patients underwent IBR (76% cancer versus 24% RR). Implant-based IBR rate was similar in both groups (58% cancer versus 63% RR). Reconstruction loss (5.3% cancer versus 4.2% RR) and complication (16% cancer versus 12.9% RR) rates were similar. Cancer patients were more likely to undergo secondary surgery (68.4% versus 56.3%; P = 0.025), including contralateral symmetrization (22.8% versus 0%) and conversion to autologous reconstruction (5.7% versus 1.4%). Secondary surgeries were mostly planned for cancer patients (72% planned versus 28% unplanned), with rates unaffected by adjuvant therapies. This distribution was different in RR patients (51.3% planned versus 48.7% unplanned). The commonest secondary procedure was lipomodeling (19.7% cancer versus 23.9% RR). For cancer patients, complications resulted in a significantly higher unplanned secondary surgery rate (82.5% versus 38.8%; P = 0.001) than patients without complications. This was not evident in the RR patients, where complications did not lead to a significantly higher unplanned surgery rate (58.9% versus 35.2%; P = 0.086). Conclusions:. Most of the secondary surgeries were planned for cancer patients. However, complications led to a significantly higher rate of unplanned secondary surgery. Approximately 1 in 4 RR patients received unplanned secondary surgery, which may be driven by the desire to achieve an optimal aesthetic outcome.

Published
2020
Full Text
View/download PDF

21. Sparse MTTKRP Acceleration for Tensor Decomposition on GPU

Author

Wijeratne, Sasindu, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Hardware Architecture
Abstract

Sparse Matricized Tensor Times Khatri-Rao Product (spMTTKRP) is the bottleneck kernel of sparse tensor decomposition. In this work, we propose a GPU-based algorithm design to address the key challenges in accelerating spMTTKRP computation, including (1) eliminating global atomic operations across GPU thread blocks, (2) avoiding the intermediate values being communicated between GPU thread blocks and GPU global memory, and (3) ensuring a balanced distribution of workloads across GPU thread blocks. Our approach also supports dynamic tensor remapping, enabling the above optimizations in all the modes of the input tensor. Our approach achieves a geometric mean speedup of 1.5x, 2.0x, and 21.7x in total execution time across widely used datasets compared with the state-of-the-art GPU implementations. Our work is the only GPU implementation that can support tensors with modes greater than 4 since the state-of-the-art works have implementation constraints for tensors with a large number of modes., Comment: In 21st ACM International Conference on Computing Frontiers (CF '24), May 7-9, 2024, Ischia, Italy

Published
2024
Full Text
View/download PDF

22. On the Power of Interactive Proofs for Learning

Author

Gur, Tom, Jahanara, Mohammad Mahdi, Khodabandeh, Mohammad Mahdi, Rajgopal, Ninad, Salamatian, Bahar, and Shinkar, Igor

Subjects
Computer Science - Computational Complexity, Computer Science - Data Structures and Algorithms, Computer Science - Machine Learning
Abstract

We continue the study of doubly-efficient proof systems for verifying agnostic PAC learning, for which we obtain the following results. - We construct an interactive protocol for learning the $t$ largest Fourier characters of a given function $f \colon \{0,1\}^n \to \{0,1\}$ up to an arbitrarily small error, wherein the verifier uses $\mathsf{poly}(t)$ random examples. This improves upon the Interactive Goldreich-Levin protocol of Goldwasser, Rothblum, Shafer, and Yehudayoff (ITCS 2021) whose sample complexity is $\mathsf{poly}(t,n)$. - For agnostically learning the class $\mathsf{AC}^0[2]$ under the uniform distribution, we build on the work of Carmosino, Impagliazzo, Kabanets, and Kolokolova (APPROX/RANDOM 2017) and design an interactive protocol, where given a function $f \colon \{0,1\}^n \to \{0,1\}$, the verifier learns the closest hypothesis up to $\mathsf{polylog}(n)$ multiplicative factor, using quasi-polynomially many random examples. In contrast, this class has been notoriously resistant even for constructing realisable learners (without a prover) using random examples. - For agnostically learning $k$-juntas under the uniform distribution, we obtain an interactive protocol, where the verifier uses $O(2^k)$ random examples to a given function $f \colon \{0,1\}^n \to \{0,1\}$. Crucially, the sample complexity of the verifier is independent of $n$. We also show that if we do not insist on doubly-efficient proof systems, then the model becomes trivial. Specifically, we show a protocol for an arbitrary class $\mathcal{C}$ of Boolean functions in the distribution-free setting, where the verifier uses $O(1)$ labeled examples to learn $f$., Comment: 58 pages, To appear in STOC 2024

Published
2024
Full Text
View/download PDF

23. GCV-Turbo: End-to-end Acceleration of GNN-based Computer Vision Tasks on FPGA

Author

Zhang, Bingyi, Kannan, Rajgopal, Busart, Carl, and Prasanna, Viktor

Subjects
Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Computer Vision and Pattern Recognition, Electrical Engineering and Systems Science - Image and Video Processing
Abstract

Graph neural networks (GNNs) have recently empowered various novel computer vision (CV) tasks. In GNN-based CV tasks, a combination of CNN layers and GNN layers or only GNN layers are employed. This paper introduces GCV-Turbo, a domain-specific accelerator on FPGA for end-to-end acceleration of GNN-based CV tasks. GCV-Turbo consists of two key components: (1) a \emph{novel} hardware architecture optimized for the computation kernels in both CNNs and GNNs using the same set of computation resources. (2) a PyTorch-compatible compiler that takes a user-defined model as input, performs end-to-end optimization for the computation graph of a given GNN-based CV task, and produces optimized code for hardware execution. The hardware architecture and the compiler work synergistically to support a variety of GNN-based CV tasks. We implement GCV-Turbo on a state-of-the-art FPGA and evaluate its performance across six representative GNN-based CV tasks with diverse input data modalities (e.g., image, human skeleton, point cloud). Compared with state-of-the-art CPU (GPU) implementations, GCV-Turbo achieves an average latency reduction of $68.4\times$ ($4.1\times$) on these six GNN-based CV tasks. Moreover, GCV-Turbo supports the execution of the standalone CNNs or GNNs, achieving performance comparable to that of state-of-the-art CNN (GNN) accelerators for widely used CNN-only (GNN-only) models.

Published
2024

24. VTR: An Optimized Vision Transformer for SAR ATR Acceleration on FPGA

Author

Wickramasinghe, Sachini, Parikh, Dhruv, Zhang, Bingyi, Kannan, Rajgopal, Prasanna, Viktor, and Busart, Carl

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence, Computer Science - Hardware Architecture, Computer Science - Distributed, Parallel, and Cluster Computing
Abstract

Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR) is a key technique used in military applications like remote-sensing image recognition. Vision Transformers (ViTs) are the current state-of-the-art in various computer vision applications, outperforming their CNN counterparts. However, using ViTs for SAR ATR applications is challenging due to (1) standard ViTs require extensive training data to generalize well due to their low locality; the standard SAR datasets, however, have a limited number of labeled training data which reduces the learning capability of ViTs; (2) ViTs have a high parameter count and are computation intensive which makes their deployment on resource-constrained SAR platforms difficult. In this work, we develop a lightweight ViT model that can be trained directly on small datasets without any pre-training by utilizing the Shifted Patch Tokenization (SPT) and Locality Self-Attention (LSA) modules. We directly train this model on SAR datasets which have limited training samples to evaluate its effectiveness for SAR ATR applications. We evaluate our proposed model, that we call VTR (ViT for SAR ATR), on three widely used SAR datasets: MSTAR, SynthWakeSAR, and GBSAR. Further, we propose a novel FPGA accelerator for VTR, in order to enable deployment for real-time SAR ATR applications., Comment: SPIE DCS 2024

Published
2024

25. FACTUAL: A Novel Framework for Contrastive Learning Based Robust SAR Image Classification

Author

Wang, Xu, Ye, Tian, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Deep Learning (DL) Models for Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR), while delivering improved performance, have been shown to be quite vulnerable to adversarial attacks. Existing works improve robustness by training models on adversarial samples. However, by focusing mostly on attacks that manipulate images randomly, they neglect the real-world feasibility of such attacks. In this paper, we propose FACTUAL, a novel Contrastive Learning framework for Adversarial Training and robust SAR classification. FACTUAL consists of two components: (1) Differing from existing works, a novel perturbation scheme that incorporates realistic physical adversarial attacks (such as OTSA) to build a supervised adversarial pre-training network. This network utilizes class labels for clustering clean and perturbed images together into a more informative feature space. (2) A linear classifier cascaded after the encoder to use the computed representations to predict the target labels. By pre-training and fine-tuning our model on both clean and adversarial samples, we show that our model achieves high prediction accuracy on both cases. Our model achieves 99.7% accuracy on clean samples, and 89.6% on perturbed samples, both outperforming previous state-of-the-art methods., Comment: 2024 IEEE Radar Conference

Published
2024

26. Uncertainty-Aware SAR ATR: Defending Against Adversarial Attacks via Bayesian Neural Networks

Author

Ye, Tian, Kannan, Rajgopal, Prasanna, Viktor, and Busart, Carl

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Adversarial attacks have demonstrated the vulnerability of Machine Learning (ML) image classifiers in Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR) systems. An adversarial attack can deceive the classifier into making incorrect predictions by perturbing the input SAR images, for example, with a few scatterers attached to the on-ground objects. Therefore, it is critical to develop robust SAR ATR systems that can detect potential adversarial attacks by leveraging the inherent uncertainty in ML classifiers, thereby effectively alerting human decision-makers. In this paper, we propose a novel uncertainty-aware SAR ATR for detecting adversarial attacks. Specifically, we leverage the capability of Bayesian Neural Networks (BNNs) in performing image classification with quantified epistemic uncertainty to measure the confidence for each input SAR image. By evaluating the uncertainty, our method alerts when the input SAR image is likely to be adversarially generated. Simultaneously, we also generate visual explanations that reveal the specific regions in the SAR image where the adversarial scatterers are likely to to be present, thus aiding human decision-making with hints of evidence of adversarial attacks. Experiments on the MSTAR dataset demonstrate that our approach can identify over 80% adversarial SAR images with fewer than 20% false alarms, and our visual explanations can identify up to over 90% of scatterers in an adversarial SAR image.

Published
2024

27. Accelerating ViT Inference on FPGA through Static and Dynamic Pruning

Author

Parikh, Dhruv, Li, Shouyi, Zhang, Bingyi, Kannan, Rajgopal, Busart, Carl, and Prasanna, Viktor

Subjects
Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Hardware Architecture, Computer Science - Computer Vision and Pattern Recognition
Abstract

Vision Transformers (ViTs) have achieved state-of-the-art accuracy on various computer vision tasks. However, their high computational complexity prevents them from being applied to many real-world applications. Weight and token pruning are two well-known methods for reducing complexity: weight pruning reduces the model size and associated computational demands, while token pruning further dynamically reduces the computation based on the input. Combining these two techniques should significantly reduce computation complexity and model size; however, naively integrating them results in irregular computation patterns, leading to significant accuracy drops and difficulties in hardware acceleration. Addressing the above challenges, we propose a comprehensive algorithm-hardware codesign for accelerating ViT on FPGA through simultaneous pruning -combining static weight pruning and dynamic token pruning. For algorithm design, we systematically combine a hardware-aware structured block-pruning method for pruning model parameters and a dynamic token pruning method for removing unimportant token vectors. Moreover, we design a novel training algorithm to recover the model's accuracy. For hardware design, we develop a novel hardware accelerator for executing the pruned model. The proposed hardware design employs multi-level parallelism with load balancing strategy to efficiently deal with the irregular computation pattern led by the two pruning approaches. Moreover, we develop an efficient hardware mechanism for efficiently executing the on-the-fly token pruning., Comment: FCCM 2024

Published
2024

28. A Collision Cone Approach for Control Barrier Functions

Author

Tayal, Manan, Goswami, Bhavya Giri, Rajgopal, Karthik, Singh, Rajpal, Rao, Tejas, Keshavan, Jishnu, Jagtap, Pushpak, and Kolathaya, Shishir

Subjects
Computer Science - Robotics
Abstract

This work presents a unified approach for collision avoidance using Collision-Cone Control Barrier Functions (CBFs) in both ground (UGV) and aerial (UAV) unmanned vehicles. We propose a novel CBF formulation inspired by collision cones, to ensure safety by constraining the relative velocity between the vehicle and the obstacle to always point away from each other. The efficacy of this approach is demonstrated through simulations and hardware implementations on the TurtleBot, Stoch-Jeep, and Crazyflie 2.1 quadrotor robot, showcasing its effectiveness in avoiding collisions with dynamic obstacles in both ground and aerial settings. The real-time controller is developed using CBF Quadratic Programs (CBF-QPs). Comparative analysis with the state-of-the-art CBFs highlights the less conservative nature of the proposed approach. Overall, this research contributes to a novel control formation that can give a guarantee for collision avoidance in unmanned vehicles by modifying the control inputs from existing path-planning controllers., Comment: 13 pages, 16 pages. arXiv admin note: substantial text overlap with arXiv:2209.11524, arXiv:2303.15871, arXiv:2310.10839

Published
2024

31. PaCKD: Pattern-Clustered Knowledge Distillation for Compressing Memory Access Prediction Models

Author

Gupta, Neelesh, Zhang, Pengmiao, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning, Computer Science - Hardware Architecture
Abstract

Deep neural networks (DNNs) have proven to be effective models for accurate Memory Access Prediction (MAP), a critical task in mitigating memory latency through data prefetching. However, existing DNN-based MAP models suffer from the challenges such as significant physical storage space and poor inference latency, primarily due to their large number of parameters. These limitations render them impractical for deployment in real-world scenarios. In this paper, we propose PaCKD, a Pattern-Clustered Knowledge Distillation approach to compress MAP models while maintaining the prediction performance. The PaCKD approach encompasses three steps: clustering memory access sequences into distinct partitions involving similar patterns, training large pattern-specific teacher models for memory access prediction for each partition, and training a single lightweight student model by distilling the knowledge from the trained pattern-specific teachers. We evaluate our approach on LSTM, MLP-Mixer, and ResNet models, as they exhibit diverse structures and are widely used for image classification tasks in order to test their effectiveness in four widely used graph applications. Compared to the teacher models with 5.406M parameters and an F1-score of 0.4626, our student models achieve a 552$\times$ model size compression while maintaining an F1-score of 0.4538 (with a 1.92% performance drop). Our approach yields an 8.70% higher result compared to student models trained with standard knowledge distillation and an 8.88% higher result compared to student models trained without any form of knowledge distillation., Comment: 6 pages, 2 figures, HPEC '23

Published
2024
Full Text
View/download PDF

32. TASER: Temporal Adaptive Sampling for Fast and Accurate Dynamic Graph Representation Learning

Author

Deng, Gangda, Zhou, Hongkuan, Zeng, Hanqing, Xia, Yinglong, Leung, Christopher, Li, Jianbo, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

Recently, Temporal Graph Neural Networks (TGNNs) have demonstrated state-of-the-art performance in various high-impact applications, including fraud detection and content recommendation. Despite the success of TGNNs, they are prone to the prevalent noise found in real-world dynamic graphs like time-deprecated links and skewed interaction distribution. The noise causes two critical issues that significantly compromise the accuracy of TGNNs: (1) models are supervised by inferior interactions, and (2) noisy input induces high variance in the aggregated messages. However, current TGNN denoising techniques do not consider the diverse and dynamic noise pattern of each node. In addition, they also suffer from the excessive mini-batch generation overheads caused by traversing more neighbors. We believe the remedy for fast and accurate TGNNs lies in temporal adaptive sampling. In this work, we propose TASER, the first adaptive sampling method for TGNNs optimized for accuracy, efficiency, and scalability. TASER adapts its mini-batch selection based on training dynamics and temporal neighbor selection based on the contextual, structural, and temporal properties of past interactions. To alleviate the bottleneck in mini-batch generation, TASER implements a pure GPU-based temporal neighbor finder and a dedicated GPU feature cache. We evaluate the performance of TASER using two state-of-the-art backbone TGNNs. On five popular datasets, TASER outperforms the corresponding baselines by an average of 2.3% in Mean Reciprocal Rank (MRR) while achieving an average of 5.1x speedup in training time., Comment: IPDPS 2024

Published
2024

33. A Single Graph Convolution Is All You Need: Efficient Grayscale Image Classification

Author

Fein-Ashley, Jacob, Wickramasinghe, Sachini, Zhang, Bingyi, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Image classifiers for domain-specific tasks like Synthetic Aperture Radar Automatic Target Recognition (SAR ATR) and chest X-ray classification often rely on convolutional neural networks (CNNs). These networks, while powerful, experience high latency due to the number of operations they perform, which can be problematic in real-time applications. Many image classification models are designed to work with both RGB and grayscale datasets, but classifiers that operate solely on grayscale images are less common. Grayscale image classification has critical applications in fields such as medical imaging and SAR ATR. In response, we present a novel grayscale image classification approach using a vectorized view of images. By leveraging the lightweight nature of Multi-Layer Perceptrons (MLPs), we treat images as vectors, simplifying the problem to grayscale image classification. Our approach incorporates a single graph convolutional layer in a batch-wise manner, enhancing accuracy and reducing performance variance. Additionally, we develop a customized accelerator on FPGA for our model, incorporating several optimizations to improve performance. Experimental results on benchmark grayscale image datasets demonstrate the effectiveness of our approach, achieving significantly lower latency (up to $16\times$ less on MSTAR) and competitive or superior performance compared to state-of-the-art models for SAR ATR and medical image classification., Comment: Accepted to IEEE ICIP 2024

Published
2024

34. PAHD: Perception-Action based Human Decision Making using Explainable Graph Neural Networks on SAR Images

Author

Wijeratne, Sasindu, Zhang, Bingyi, Kannan, Rajgopal, Prasanna, Viktor, and Busart, Carl

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning, Electrical Engineering and Systems Science - Image and Video Processing
Abstract

Synthetic Aperture Radar (SAR) images are commonly utilized in military applications for automatic target recognition (ATR). Machine learning (ML) methods, such as Convolutional Neural Networks (CNN) and Graph Neural Networks (GNN), are frequently used to identify ground-based objects, including battle tanks, personnel carriers, and missile launchers. Determining the vehicle class, such as the BRDM2 tank, BMP2 tank, BTR60 tank, and BTR70 tank, is crucial, as it can help determine whether the target object is an ally or an enemy. While the ML algorithm provides feedback on the recognized target, the final decision is left to the commanding officers. Therefore, providing detailed information alongside the identified target can significantly impact their actions. This detailed information includes the SAR image features that contributed to the classification, the classification confidence, and the probability of the identified object being classified as a different object type or class. We propose a GNN-based ATR framework that provides the final classified class and outputs the detailed information mentioned above. This is the first study to provide a detailed analysis of the classification class, making final decisions more straightforward. Moreover, our GNN framework achieves an overall accuracy of 99.2\% when evaluated on the MSTAR dataset, improving over previous state-of-the-art GNN methods.

Published
2024

35. Attention, Distillation, and Tabularization: Towards Practical Neural Network-Based Prefetching

Author

Zhang, Pengmiao, Gupta, Neelesh, Kannan, Rajgopal, and Prasanna, Viktor K.

Subjects
Computer Science - Neural and Evolutionary Computing, Computer Science - Hardware Architecture, Computer Science - Machine Learning, Computer Science - Operating Systems
Abstract

Attention-based Neural Networks (NN) have demonstrated their effectiveness in accurate memory access prediction, an essential step in data prefetching. However, the substantial computational overheads associated with these models result in high inference latency, limiting their feasibility as practical prefetchers. To close the gap, we propose a new approach based on tabularization that significantly reduces model complexity and inference latency without sacrificing prediction accuracy. Our novel tabularization methodology takes as input a distilled, yet highly accurate attention-based model for memory access prediction and efficiently converts its expensive matrix multiplications into a hierarchy of fast table lookups. As an exemplar of the above approach, we develop DART, a prefetcher comprised of a simple hierarchy of tables. With a modest 0.09 drop in F1-score, DART reduces 99.99% of arithmetic operations from the large attention-based model and 91.83% from the distilled model. DART accelerates the large model inference by 170x and the distilled model by 9.4x. DART has comparable latency and storage costs as state-of-the-art rule-based prefetcher BO but surpasses it by 6.1% in IPC improvement. DART outperforms state-of-the-art NN-based prefetchers TransFetch by 33.1% and Voyager by 37.2% in terms of IPC improvement, primarily due to its low prefetching latency.

Published
2023

36. Stoch BiRo: Design and Control of a low cost bipedal robot

Author

Mothish, GVS, Rajgopal, Karthik, Kola, Ravi, Tayal, Manan, and Kolathaya, Shishir

Subjects
Computer Science - Robotics
Abstract

This paper introduces the Stoch BiRo, a cost-effective bipedal robot designed with a modular mechanical structure having point feet to navigate uneven and unfamiliar terrains. The robot employs proprioceptive actuation in abduction, hips, and knees, leveraging a Raspberry Pi4 for control. Overcoming computational limitations, a Learning-based Linear Policy controller manages balance and locomotion with only 3 degrees of freedom (DoF) per leg, distinct from the typical 5DoF in bipedal systems. Integrated within a modular control architecture, these controllers enable autonomous handling of unforeseen terrain disturbances without external sensors or prior environment knowledge. The robot's policies are trained and simulated using MuJoCo, transferring learned behaviors to the Stoch BiRo hardware for initial walking validations. This work highlights the Stoch BiRo's adaptability and cost-effectiveness in mechanical design, control strategies, and autonomous navigation, promising diverse applications in real-world robotics scenarios., Comment: 7 Pages, 6 figures

Published
2023

37. Benchmarking Deep Learning Classifiers for SAR Automatic Target Recognition

Author

Fein-Ashley, Jacob, Ye, Tian, Kannan, Rajgopal, Prasanna, Viktor, and Busart, Carl

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning, Electrical Engineering and Systems Science - Image and Video Processing
Abstract

Synthetic Aperture Radar SAR Automatic Target Recognition ATR is a key technique of remote-sensing image recognition which can be supported by deep neural networks The existing works of SAR ATR mostly focus on improving the accuracy of the target recognition while ignoring the systems performance in terms of speed and storage which is critical to real-world applications of SAR ATR For decision-makers aiming to identify a proper deep learning model to deploy in a SAR ATR system it is important to understand the performance of different candidate deep learning models and determine the best model accordingly This paper comprehensively benchmarks several advanced deep learning models for SAR ATR with multiple distinct SAR imagery datasets Specifically we train and test five SAR image classifiers based on Residual Neural Networks ResNet18 ResNet34 ResNet50 Graph Neural Network GNN and Vision Transformer for Small-Sized Datasets (SS-ViT) We select three datasets MSTAR GBSAR and SynthWakeSAR that offer heterogeneity We evaluate and compare the five classifiers concerning their classification accuracy runtime performance in terms of inference throughput and analytical performance in terms of number of parameters number of layers model size and number of operations Experimental results show that the GNN classifier outperforms with respect to throughput and latency However it is also shown that no clear model winner emerges from all of our chosen metrics and a one model rules all case is doubtful in the domain of SAR ATR, Comment: 6 Pages

Published
2023
Full Text
View/download PDF

38. Realistic Scatterer Based Adversarial Attacks on SAR Image Classifiers

Author

Ye, Tian, Kannan, Rajgopal, Prasanna, Viktor, Busart, Carl, and Kaplan, Lance

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Adversarial attacks have highlighted the vulnerability of classifiers based on machine learning for Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR) tasks. An adversarial attack perturbs SAR images of on-ground targets such that the classifiers are misled into making incorrect predictions. However, many existing attacking techniques rely on arbitrary manipulation of SAR images while overlooking the feasibility of executing the attacks on real-world SAR imagery. Instead, adversarial attacks should be able to be implemented by physical actions, for example, placing additional false objects as scatterers around the on-ground target to perturb the SAR image and fool the SAR ATR. In this paper, we propose the On-Target Scatterer Attack (OTSA), a scatterer-based physical adversarial attack. To ensure the feasibility of its physical execution, we enforce a constraint on the positioning of the scatterers. Specifically, we restrict the scatterers to be placed only on the target instead of in the shadow regions or the background. To achieve this, we introduce a positioning score based on Gaussian kernels and formulate an optimization problem for our OTSA attack. Using a gradient ascent method to solve the optimization problem, the OTSA can generate a vector of parameters describing the positions, shapes, sizes and amplitudes of the scatterers to guide the physical execution of the attack that will mislead SAR image classifiers. The experimental results show that our attack obtains significantly higher success rates under the positioning constraint compared with the existing method.

Published
2023

39. A randomized controlled trial to compare the effects of time-restricted eating versus Mediterranean diet on symptoms and quality of life in bipolar disorder.

Author

Murray, Greg, Kriegsfeld, Lance, Manoogian, Emily, Mason, Liam, Allen, J, Berk, Michael, Panda, Satchidanda, Rajgopal, Nandini, Gibson, Jake, Joyner, Keanan, Villanueva, Robert, Michalak, Erin, and Johnson, Sheri

Subjects
Bipolar disorder, Circadian, Depression, Mania, Mediterranean diet, Quality of life, Randomized controlled trial, Time-restricted eating, Humans, Bipolar Disorder, Diet, Mediterranean, Quality of Life, Sleep Wake Disorders, Adult, Female, Male, Circadian Rhythm
Abstract

BACKGROUND: The primary objective of this randomized controlled trial (RCT) is to establish the effectiveness of time-restricted eating (TRE) compared with the Mediterranean diet for people with bipolar disorder (BD) who have symptoms of sleep disorders or circadian rhythm sleep-wake disruption. This work builds on the growing evidence that TRE has benefits for improving circadian rhythms. TRE and Mediterranean diet guidance will be offered remotely using self-help materials and an app, with coaching support. METHODS: This study is an international RCT to compare the effectiveness of TRE and the Mediterranean diet. Three hundred participants will be recruited primarily via social media. Main inclusion criteria are: receiving treatment for a diagnosis of BD I or II (confirmed via DIAMOND structured diagnostic interview), endorsement of sleep or circadian problems, self-reported eating window of ≥ 12 h, and no current mood episode, acute suicidality, eating disorder, psychosis, alcohol or substance use disorder, or other health conditions that would interfere with or limit the safety of following the dietary guidance. Participants will be asked to complete baseline daily food logging for two weeks and then will be randomly allocated to follow TRE or the Mediterranean diet for 8 weeks, during which time, they will continue to complete daily food logging. Intervention content will be delivered via an app. Symptom severity interviews will be conducted at baseline; mid-intervention (4 weeks after the intervention begins); end of intervention; and at 6, 9, and 15 months post-baseline by phone or videoconference. Self-rated symptom severity and quality of life data will be gathered at those timepoints, as well as at 16 weeks post baseline. To provide a more refined index of whether TRE successfully decreases emotional lability and improves sleep, participants will be asked to complete a sleep diary (core CSD) each morning and complete six mood assessments per day for eight days at baseline and again at mid-intervention. DISCUSSION: The planned research will provide novel and important information on whether TRE is more beneficial than the Mediterranean diet for reducing mood symptoms and improving quality of life in individuals with BD who also experience sleep or circadian problems. TRIAL REGISTRATION: ClinicalTrials.gov ID NCT06188754.

Published
2024

42. Critical assessment of variant prioritization methods for rare disease diagnosis within the rare genomes project

Author

Stenton, Sarah L., O’Leary, Melanie C., Lemire, Gabrielle, VanNoy, Grace E., DiTroia, Stephanie, Ganesh, Vijay S., Groopman, Emily, O’Heir, Emily, Mangilog, Brian, Osei-Owusu, Ikeoluwa, Pais, Lynn S., Serrano, Jillian, Singer-Berk, Moriel, Weisburd, Ben, Wilson, Michael W., Austin-Tse, Christina, Abdelhakim, Marwa, Althagafi, Azza, Babbi, Giulia, Bellazzi, Riccardo, Bovo, Samuele, Carta, Maria Giulia, Casadio, Rita, Coenen, Pieter-Jan, De Paoli, Federica, Floris, Matteo, Gajapathy, Manavalan, Hoehndorf, Robert, Jacobsen, Julius O. B., Joseph, Thomas, Kamandula, Akash, Katsonis, Panagiotis, Kint, Cyrielle, Lichtarge, Olivier, Limongelli, Ivan, Lu, Yulan, Magni, Paolo, Mamidi, Tarun Karthik Kumar, Martelli, Pier Luigi, Mulargia, Marta, Nicora, Giovanna, Nykamp, Keith, Pejaver, Vikas, Peng, Yisu, Pham, Thi Hong Cam, Podda, Maurizio S., Rao, Aditya, Rizzo, Ettore, Saipradeep, Vangala G., Savojardo, Castrense, Schols, Peter, Shen, Yang, Sivadasan, Naveen, Smedley, Damian, Soru, Dorian, Srinivasan, Rajgopal, Sun, Yuanfei, Sunderam, Uma, Tan, Wuwei, Tiwari, Naina, Wang, Xiao, Wang, Yaqiong, Williams, Amanda, Worthey, Elizabeth A., Yin, Rujie, You, Yuning, Zeiberg, Daniel, Zucca, Susanna, Bakolitsa, Constantina, Brenner, Steven E., Fullerton, Stephanie M., Radivojac, Predrag, Rehm, Heidi L., and O’Donnell-Luria, Anne

Published
2024
Full Text
View/download PDF

43. Reuse Kernels or Activations? A Flexible Dataflow for Low-latency Spectral CNN Acceleration

Author

Niu, Yue, Kannan, Rajgopal, Srivastava, Ajitesh, and Prasanna, Viktor

Subjects
Computer Science - Hardware Architecture, Electrical Engineering and Systems Science - Signal Processing
Abstract

Spectral-domain CNNs have been shown to be more efficient than traditional spatial CNNs in terms of reducing computation complexity. However they come with a `kernel explosion' problem that, even after compression (pruning), imposes a high memory burden and off-chip bandwidth requirement for kernel access. This creates a performance gap between the potential acceleration offered by compression and actual FPGA implementation performance, especially for low-latency CNN inference. In this paper, we develop a principled approach to overcoming this performance gap and designing a low-latency, low-bandwidth, spectral sparse CNN accelerator on FPGAs. First, we analyze the bandwidth-storage tradeoff of sparse convolutional layers and locate communication bottlenecks. We then develop a dataflow for flexibly optimizing data reuse in different layers to minimize off-chip communication. Finally, we propose a novel scheduling algorithm to optimally schedule the on-chip memory access of multiple sparse kernels and minimize read conflicts. On a state-of-the-art FPGA platform, our design reduces data transfers by 42\% with DSP utilization up to 90\% and achieves inference latency of 9 ms for VGG16, compared to the baseline state-of-the-art latency of 68 ms., Comment: 11 pages, 11 figures Accepted to ACM/SIGDA International Symposium on Field-Programmable Gate Arrays (FPGA) 2020

Published
2023

44. Collision Cone Control Barrier Functions: Experimental Validation on UGVs for Kinematic Obstacle Avoidance

Author

Goswami, Bhavya Giri, Tayal, Manan, Rajgopal, Karthik, Jagtap, Pushpak, and Kolathaya, Shishir

Subjects
Computer Science - Robotics, Electrical Engineering and Systems Science - Systems and Control, Mathematics - Optimization and Control
Abstract

Autonomy advances have enabled robots in diverse environments and close human interaction, necessitating controllers with formal safety guarantees. This paper introduces an experimental platform designed for the validation and demonstration of a novel class of Control Barrier Functions (CBFs) tailored for Unmanned Ground Vehicles (UGVs) to proactively prevent collisions with kinematic obstacles by integrating the concept of collision cones. While existing CBF formulations excel with static obstacles, extensions to torque/acceleration-controlled unicycle and bicycle models have seen limited success. Conventional CBF applications in nonholonomic UGV models have demonstrated control conservatism, particularly in scenarios where steering/thrust control was deemed infeasible. Drawing inspiration from collision cones in path planning, we present a pioneering CBF formulation ensuring theoretical safety guarantees for both unicycle and bicycle models. The core premise revolves around aligning the obstacle's velocity away from the vehicle, establishing a constraint to perpetually avoid vectors directed towards it. This control methodology is rigorously validated through simulations and experimental verification on the Copernicus mobile robot (Unicycle Model) and FOCAS-Car (Bicycle Model)., Comment: 8 pages, 11 figures, Submitted at American Control Conference (ACC), 2024. arXiv admin note: substantial text overlap with arXiv:2209.11524

Published
2023

45. Performance of Graph Neural Networks for Point Cloud Applications

Author

Parikh, Dhruv, Zhang, Bingyi, Kannan, Rajgopal, Prasanna, Viktor, and Busart, Carl

Subjects
Computer Science - Distributed, Parallel, and Cluster Computing
Abstract

Graph Neural Networks (GNNs) have gained significant momentum recently due to their capability to learn on unstructured graph data. Dynamic GNNs (DGNNs) are the current state-of-the-art for point cloud applications; such applications (viz. autonomous driving) require real-time processing at the edge with tight latency and memory constraints. Conducting performance analysis on such DGNNs, thus, becomes a crucial task to evaluate network suitability. This paper presents a profiling analysis of EdgeConv-based DGNNs applied to point cloud inputs. We assess their inference performance in terms of end-to-end latency and memory consumption on state-of-the-art CPU and GPU platforms. The EdgeConv layer has two stages: (1) dynamic graph generation using k-Nearest Neighbors (kNN) and, (2) node feature updation. The addition of dynamic graph generation via kNN in each (EdgeConv) layer enhances network performance compared to networks that work with the same static graph in each layer; such performance enhancement comes, however, at the added computational cost associated with the dynamic graph generation stage (via kNN algorithm). Understanding its costs is essential for identifying the performance bottleneck and exploring potential avenues for hardware acceleration. To this end, this paper aims to shed light on the performance characteristics of EdgeConv-based DGNNs for point cloud inputs. Our performance analysis on a state-of-the-art EdgeConv network for classification shows that the dynamic graph construction via kNN takes up upwards of 95% of network latency on the GPU and almost 90% on the CPU. Moreover, we propose a quasi-Dynamic Graph Neural Network (qDGNN) that halts dynamic graph updates after a specific depth within the network to significantly reduce the latency on both CPU and GPU whilst matching the original networks inference accuracy., Comment: 27th Annual IEEE High Performance Extreme Computing Conference

Published
2023

46. Dynasor: A Dynamic Memory Layout for Accelerating Sparse MTTKRP for Tensor Decomposition on Multi-core CPU

Author

Wijeratne, Sasindu, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Distributed, Parallel, and Cluster Computing
Abstract

Sparse Matricized Tensor Times Khatri-Rao Product (spMTTKRP) is the most time-consuming compute kernel in sparse tensor decomposition. In this paper, we introduce a novel algorithm to minimize the execution time of spMTTKRP across all modes of an input tensor on multi-core CPU platform. The proposed algorithm leverages the FLYCOO tensor format to exploit data locality in external memory accesses. It effectively utilizes computational resources by enabling lock-free concurrent processing of independent partitions of the input tensor. The proposed partitioning ensures load balancing among CPU threads. Our dynamic tensor remapping technique leads to reduced communication overhead along all the modes. On widely used real-world tensors, our work achieves 2.12x - 9.01x speedup in total execution time across all modes compared with the state-of-the-art CPU implementations.

Published
2023

47. Characterizing Speed Performance of Multi-Agent Reinforcement Learning

Author

Wiggins, Samuel, Meng, Yuan, Kannan, Rajgopal, and Prasanna, Viktor

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Multiagent Systems
Abstract

Multi-Agent Reinforcement Learning (MARL) has achieved significant success in large-scale AI systems and big-data applications such as smart grids, surveillance, etc. Existing advancements in MARL algorithms focus on improving the rewards obtained by introducing various mechanisms for inter-agent cooperation. However, these optimizations are usually compute- and memory-intensive, thus leading to suboptimal speed performance in end-to-end training time. In this work, we analyze the speed performance (i.e., latency-bounded throughput) as the key metric in MARL implementations. Specifically, we first introduce a taxonomy of MARL algorithms from an acceleration perspective categorized by (1) training scheme and (2) communication method. Using our taxonomy, we identify three state-of-the-art MARL algorithms - Multi-Agent Deep Deterministic Policy Gradient (MADDPG), Target-oriented Multi-agent Communication and Cooperation (ToM2C), and Networked Multi-Agent RL (NeurComm) - as target benchmark algorithms, and provide a systematic analysis of their performance bottlenecks on a homogeneous multi-core CPU platform. We justify the need for MARL latency-bounded throughput to be a key performance metric in future literature while also addressing opportunities for parallelization and acceleration.

Published
2023
Full Text
View/download PDF

48. CAGI, the Critical Assessment of Genome Interpretation, establishes progress and prospects for computational genetic variant interpretation methods

Author

Jain, Shantanu, Bakolitsa, Constantina, Brenner, Steven E, Radivojac, Predrag, Moult, John, Repo, Susanna, Hoskins, Roger A, Andreoletti, Gaia, Barsky, Daniel, Chellapan, Ajithavalli, Chu, Hoyin, Dabbiru, Navya, Kollipara, Naveen K, Ly, Melissa, Neumann, Andrew J, Pal, Lipika R, Odell, Eric, Pandey, Gaurav, Peters-Petrulewicz, Robin C, Srinivasan, Rajgopal, Yee, Stephen F, Yeleswarapu, Sri Jyothsna, Zuhl, Maya, Adebali, Ogun, Patra, Ayoti, Beer, Michael A, Hosur, Raghavendra, Peng, Jian, Bernard, Brady M, Berry, Michael, Dong, Shengcheng, Boyle, Alan P, Adhikari, Aashish, Chen, Jingqi, Hu, Zhiqiang, Wang, Robert, Wang, Yaqiong, Miller, Maximilian, Wang, Yanran, Bromberg, Yana, Turina, Paola, Capriotti, Emidio, Han, James J, Ozturk, Kivilcim, Carter, Hannah, Babbi, Giulia, Bovo, Samuele, Di Lena, Pietro, Martelli, Pier Luigi, Savojardo, Castrense, Casadio, Rita, Cline, Melissa S, De Baets, Greet, Bonache, Sandra, Díez, Orland, Gutiérrez-Enríquez, Sara, Fernández, Alejandro, Montalban, Gemma, Ootes, Lars, Özkan, Selen, Padilla, Natàlia, Riera, Casandra, De la Cruz, Xavier, Diekhans, Mark, Huwe, Peter J, Wei, Qiong, Xu, Qifang, Dunbrack, Roland L, Gotea, Valer, Elnitski, Laura, Margolin, Gennady, Fariselli, Piero, Kulakovskiy, Ivan V, Makeev, Vsevolod J, Penzar, Dmitry D, Vorontsov, Ilya E, Favorov, Alexander V, Forman, Julia R, Hasenahuer, Marcia, Fornasari, Maria S, Parisi, Gustavo, Avsec, Ziga, Çelik, Muhammed H, Nguyen, Thi Yen Duong, Gagneur, Julien, Shi, Fang-Yuan, Edwards, Matthew D, Guo, Yuchun, Tian, Kevin, Zeng, Haoyang, Gifford, David K, Göke, Jonathan, Zaucha, Jan, Gough, Julian, Ritchie, Graham RS, Frankish, Adam, Mudge, Jonathan M, Harrow, Jennifer, Young, Erin L, and Yu, Yao

Subjects
Biological Sciences, Genetics, Human Genome, 2.1 Biological and endogenous factors, Good Health and Well Being, Humans, Computational Biology, Mutation, Missense, Phenotype, Critical Assessment of Genome Interpretation Consortium, Environmental Sciences, Information and Computing Sciences, Bioinformatics
Abstract

BackgroundThe Critical Assessment of Genome Interpretation (CAGI) aims to advance the state-of-the-art for computational prediction of genetic variant impact, particularly where relevant to disease. The five complete editions of the CAGI community experiment comprised 50 challenges, in which participants made blind predictions of phenotypes from genetic data, and these were evaluated by independent assessors.ResultsPerformance was particularly strong for clinical pathogenic variants, including some difficult-to-diagnose cases, and extends to interpretation of cancer-related variants. Missense variant interpretation methods were able to estimate biochemical effects with increasing accuracy. Assessment of methods for regulatory variants and complex trait disease risk was less definitive and indicates performance potentially suitable for auxiliary use in the clinic.ConclusionsResults show that while current methods are imperfect, they have major utility for research and clinical applications. Emerging methods and increasingly large, robust datasets for training and assessment promise further progress ahead.

Published
2024

50. Distribution-Free Proofs of Proximity

Author

Aaronson, Hugo, Gur, Tom, Rajgopal, Ninad, and Rothblum, Ron D.

Subjects
Computer Science - Computational Complexity
Abstract

Motivated by the fact that input distributions are often unknown in advance, distribution-free property testing considers a setting where the algorithmic task is to accept functions $f : [n] \to \{0,1\}$ with a certain property P and reject functions that are $\eta$-far from P, where the distance is measured according to an arbitrary and unknown input distribution $D \sim [n]$. As usual in property testing, the tester can only make a sublinear number of input queries, but as the distribution is unknown, we also allow a sublinear number of samples from the distribution D. In this work we initiate the study of distribution-free interactive proofs of proximity (df-IPPs) in which the distribution-free testing algorithm is assisted by an all powerful but untrusted prover. Our main result is that for any problem P $\in$ NC, any proximity parameter $\eta > 0$, and any (trade-off) parameter $t\leq\sqrt{n}$, we construct a df-IPP for P with respect to $\eta$, that has query and sample complexities $t+O(1/\eta)$, and communication complexity $\tilde{O}(n/t + 1/\eta)$. For t as above and sufficiently large $\eta$ (namely, when $\eta > t/n$), this result matches the parameters of the best-known general purpose IPPs in the standard uniform setting. Moreover, for such t, its parameters are optimal up to poly-logarithmic factors under reasonable cryptographic assumptions for the same regime of $\eta$ as the uniform setting, i.e., when $\eta \geq 1/t$. For small $\eta$ (i.e., $\eta< t/n$), our protocol has communication complexity $\Omega(1/\eta)$, which is worse than the $\tilde{O}(n/t)$ communication complexity of the uniform IPPs (with the same query complexity). To improve on this gap, we show that for IPPs over specialised, but large distribution families, such as sufficiently smooth distributions and product distributions, the communication complexity reduces to $\tilde{O}(n/t^{1-o(1)})$.

Published
2023

Catalog

Books, media, physical & digital resources
See catalog results