54,816 results on '"Wang, Yang"'
Search Results
2. Research on Coal-Rock Recognition Based on the EMD-EM-BP Model
- Author
-
Tian, Shengyu, primary, Tian, Ying, additional, Wang, Yang, additional, and Gu, Jieying, additional
- Published
- 2024
- Full Text
- View/download PDF
3. Research on Predicting Leakage Aperture of Water Pipelines Based on IGWO-BP Model
- Author
-
Tan, Tianyan, primary, Zhang, Qiang, additional, Wang, Yang, additional, and Gu, Jieying, additional
- Published
- 2024
- Full Text
- View/download PDF
4. Social media and performative parenting
- Author
-
Lim, Sun Sun, primary and Wang, Yang, additional
- Published
- 2024
- Full Text
- View/download PDF
5. Path Tracking and Speed Following Control of Four In-wheel-driving Autonomous Vehicle
- Author
-
Li, Shengfei, primary, Wang, Yang, additional, Pan, Bo, additional, Tan, Senqi, additional, Zhang, Naisi, additional, and Su, Bo, additional
- Published
- 2024
- Full Text
- View/download PDF
6. Multi-sourced Integrated Ranking with Exposure Fairness
- Author
-
Liu, Yifan, primary, Liu, Weiwen, additional, Xia, Wei, additional, Zhu, Jieming, additional, Zhang, Weinan, additional, Dong, Zhenhua, additional, Wang, Yang, additional, Tang, Ruiming, additional, Zhang, Rui, additional, and Yu, Yong, additional
- Published
- 2024
- Full Text
- View/download PDF
7. Optimization of Compensation Parameters for DC High-Voltage Power Supply Using PSO
- Author
-
Jiang, Can, primary, Lu, Yi, additional, Liu, Fangmei, additional, Wang, Yang, additional, and Deng, Fangxiong, additional
- Published
- 2024
- Full Text
- View/download PDF
8. Research on Energy Efficiency Improvement Methods for Large-Scale Air Conditioning Systems Oriented to Multiple Scenarios
- Author
-
Huang, Bowen, primary, Xiang, Jie, additional, Zeng, Zihao, additional, Wang, Yang, additional, Zhou, Yamin, additional, Liu, Shuang, additional, and Wang, Xu, additional
- Published
- 2024
- Full Text
- View/download PDF
9. Electrical Load Prediction by an Improved Long Short-Term Memory Based on Variable Dimension Reduction
- Author
-
Fu, Yu, primary, Wang, Yang, additional, Cai, Yongxiang, additional, Liu, Anjiang, additional, Wen, Yi, additional, Li, Hongwei, additional, Ren, Jiakuan, additional, and Qu, Yangquan, additional
- Published
- 2024
- Full Text
- View/download PDF
10. Approximate Analytical Calculation of Magnetic Shielding of Double-Layer Conducting Plates with Periodic Apertures
- Author
-
Huang, Jiancheng, primary, Guo, Xingxin, additional, Wang, Yang, additional, and Jiao, Chongqing, additional
- Published
- 2024
- Full Text
- View/download PDF
11. Fracture Extension Law and Process Optimization of Lianggaoshan Formation Shale Reservoir in Sichuan Basin
- Author
-
Lv, Ze-fei, primary, Li, Wen-zhe, additional, Wang, Yang, additional, Chen, Wei-hua, additional, Wang, Zhou-yang, additional, and He, Rui, additional
- Published
- 2024
- Full Text
- View/download PDF
12. Research on Adaptability of Downhole Two-Stage Series Hydrocyclone Separation Technology
- Author
-
Zhao, Yun-long, primary, Sun, Chun-long, additional, Wang, Yang, additional, Ma, Zeng-hua, additional, Wang, Tong, additional, Song, Hong-zhi, additional, Dai, Jin-ming, additional, Xing, Yun-long, additional, Gao, Jian-hua, additional, An, Hong-xin, additional, Zhang, Yang, additional, Sun, Hong-jie, additional, Liu, Ming-kai, additional, Zhen, Dong-fang, additional, Hu, Hou-meng, additional, and Gu, Qi-lin, additional
- Published
- 2024
- Full Text
- View/download PDF
13. Development and Application of a Detection System for Submersible Permanent Magnet Linear Motor
- Author
-
Du, Wei-shan, primary, Sun, Yan-an, additional, Sun, Xiu-lin, additional, Wang, Feng-ying, additional, Ge, Wei-tao, additional, Li, Ji-nan, additional, and Wang, Yang, additional
- Published
- 2024
- Full Text
- View/download PDF
14. Research and Evaluation of Variable Speed Optimization Operation Technologies for Pumping Unit in Y Oilfield
- Author
-
Wang, Yang, primary
- Published
- 2024
- Full Text
- View/download PDF
15. Analysis on Optimization of Adjusting Parameters Method for Pumping Unit in Y Block
- Author
-
Zhang, Wan-qing, primary, Hou, Yu, additional, Wang, Cui, additional, Zhang, Kai-bo, additional, Yi, Kun, additional, Wang, Yang, additional, Yin, Lei, additional, Wang, Si-qi, additional, Zhou, Lu-fang, additional, and Li, Chun-hong, additional
- Published
- 2024
- Full Text
- View/download PDF
16. A Micro-scale Method for Predicting the Mechanical Properties of Reservoirs with Strong Heterogeneity
- Author
-
He, Rui, primary, Li, Wen-zhe, additional, Chen, Wei-hua, additional, Zeng, Ji, additional, Chen, Yan, additional, Lv, Ze-fei, additional, Wang, Yang, additional, and Wang, Tao, additional
- Published
- 2024
- Full Text
- View/download PDF
17. Research and Application of Anti Eccentric Wear Technology of Rod and Tubing in Oil Production of PCP
- Author
-
Sun, Chun-long, primary, Wang, Yang, additional, Gao, Yu, additional, Yuan, Wen, additional, and Jiang, Wei, additional
- Published
- 2024
- Full Text
- View/download PDF
18. Optimizing Pointwise Convolutions on Multi-core DSPs
- Author
-
Wang, Yang, primary, Wang, Qinglin, additional, Pei, Xiangdong, additional, Mei, Songzhu, additional, and Liu, Jie, additional
- Published
- 2024
- Full Text
- View/download PDF
19. PointerNet with Local and Global Contexts for Natural Language Moment Localization
- Author
-
Ye, Linwei, primary, Liu, Zhi, additional, and Wang, Yang, additional
- Published
- 2024
- Full Text
- View/download PDF
20. Campylobacter
- Author
-
Shen, Zhangqi, primary, Wang, Yang, additional, and Shen, Jianzhong, additional
- Published
- 2024
- Full Text
- View/download PDF
21. List of contributors
- Author
-
Aizawa, Shin-Ichi, primary, Alberdi, Pilar, additional, Alexander, David C., additional, Alía, Alberto, additional, Allison, D.G., additional, Amyes, Sebastian G.B., additional, An, Haoran, additional, Andrade, María J., additional, Antelmann, Haike, additional, Arias, Cesar A., additional, Asensio, Miguel A., additional, Axell-House, Dierdre B., additional, Bae, Hee-Won, additional, Baena, Laura Muñoz, additional, Baig, Abdul Mannan, additional, Bailey, Spenser O., additional, Baize, Sylvain, additional, Baldi, Pablo C., additional, Barbosa, Angela Silva, additional, Barbuddhe, Sukhadeo B., additional, Bard, Emilie, additional, Barry, Eileen M., additional, Basarab, Gregory S., additional, Beloborodova, N.V, additional, Bermúdez, Elena, additional, Bidmos, Fadil A., additional, Bisgaard, Magne, additional, Blakely, Garry W., additional, Bloch, Evan, additional, Boesen, Thias Oberg, additional, Bose, Dipayan, additional, Botero, Javier Enrique, additional, Bouabe, Hicham, additional, Bouchard, Michael J., additional, Bozue, Joel A., additional, Bradbury, Richard S., additional, Brett Moreau, G., additional, Cabezas-Cruz, Alejandro, additional, Cai, Rong-Jun, additional, Calderón, Enrique J., additional, Cao, Boyang, additional, Carmena, David, additional, Carvalho, Eneas, additional, Caulfield, Amanda D., additional, Cen, Shan, additional, Chai, Jong-Yil, additional, Chamberland, Robin R., additional, Champredon, David, additional, Chan, Edward D., additional, Charbon, Godefroid, additional, Chato, Connor, additional, Chelomina, G.N., additional, Chen, Jingyu, additional, Chen, Min, additional, Chen, Shuyu, additional, Chen, Suilin, additional, Chen, Yanfei, additional, Chen, Zhaoyuan, additional, Cheng, Aimin, additional, Cheng, Keding, additional, Chiu, Charles Y., additional, Cho, You-Hee, additional, Christensen, Henrik, additional, Chrtdernevskaya, E.A., additional, Contreras, Adolfo, additional, Contreras, Marinela, additional, Córdoba, Juan J., additional, Córdoba, María G., additional, Costa, Rita, additional, Cote, Christopher K., additional, Cui, Xiangling, additional, Cui, Yujun, additional, Dacal, Elena, additional, Dammann, Allison N., additional, Das, Shubhagata, additional, Dashti, Alejandro, additional, de la Fuente, José, additional, de la Garza, Mireya, additional, Delgado, Josué, additional, Delgado-Cuesta, Juan, additional, Deng, Haiteng, additional, Deng, Li, additional, Dey, Debajit, additional, Dhama, Kuldeep, additional, Diego, Juan García-Bernalt, additional, Ding, Hao, additional, Doern, Christopher D., additional, Dorman, Charles J., additional, Du, Zongmin, additional, Dunbar, Sherry A., additional, Duthie, Malcolm, additional, Dybvig, Kevin F., additional, Eakin, Ann E., additional, Eallonardo, Samuel J., additional, Eberly, Allison R., additional, Echeverry, Adriana Jaramillo, additional, Egland, Paul G., additional, El Zowalaty, Mohamed E., additional, Endsley, Janice Jones, additional, Eom, Keeseon S., additional, Evans, Benjamin A., additional, Falkinham, Joseph O., additional, Feng, Siwei, additional, Feng, Yaoyu, additional, Feng, Zongdi, additional, Fernández-Soto, Pedro, additional, Ferreira, Roux-Cil, additional, Flores-Huerta, Nadia, additional, Foster, Timothy J., additional, Fox-Moon, Sandra M., additional, Fraga, Tatiana Rodrigues, additional, Fredricks, David N., additional, Freitag, Nancy E., additional, Frimodt-Møller, Jakob, additional, Fuller, Risa, additional, Ganesh, Balasubramanian, additional, Gao, Ning, additional, García-Carnero, Laura C., additional, Garzetti, Debora, additional, Geoghegan, Joan A., additional, Ghenim, Raed, additional, Giambartolomei, Guillermo H., additional, Gilbert, Nicole M., additional, Gillis, Thomas Phillip, additional, Gladstone, Camilla A., additional, Gómez-Gaviria, Manuela, additional, Gómez-Marín, Jorge E., additional, Gong, Tengfang, additional, González, Ramón A., additional, Gray-Owen, Scott D., additional, Gu, Bing, additional, Guzmán-Téllez, Paula, additional, Hajal, Caroline, additional, Han, Yanping, additional, Hao, Yi, additional, Harrington, Amanda T., additional, Harris, Jason B., additional, Harvill, Eric T., additional, Hasan, S. Saif, additional, He, Guang-Jun, additional, He, Yongqun, additional, Heffron, Jared D., additional, Hidalgo, Paloma, additional, Hindiyeh, Musa Y., additional, Hreha, Teri N., additional, Hu, Xiaoyu, additional, Huang, Guanghua, additional, Huang, Jiangqing, additional, Huang, Liang, additional, Huang, Shifeng, additional, Huang, Xingxu, additional, Huang, Xueting, additional, Huang, Yilun, additional, Huffman, Anthony, additional, Humphreys, Tricia L., additional, Hunstad, David A., additional, Inglis, Timothy J.J., additional, Isaac, Lourdes, additional, Jacobs, Samantha E., additional, Janowicz, Diane M., additional, Jeon, Hyeong-Kyu, additional, Ji, Quanjiang, additional, Jia, Qi, additional, Jia, Wei, additional, Jin, Shouguang, additional, Jneidi, Lama, additional, Jose, Shinsmon, additional, Jung, Bong-Kwang, additional, Kattan, Randa, additional, Kaushik, Rahul, additional, Khare, Reeti, additional, Kim, Eun Sook, additional, Kirn, Thomas J., additional, Koo, Hyun, additional, Köster, Pamela C., additional, Krause, Peter J., additional, Kumar, Sanjai, additional, Kupz, Andreas, additional, Lambert, P.A., additional, Lamont, Richard J., additional, Langford, Paul R., additional, Lebeaux, David, additional, Legname, Giuseppe, additional, Li, Bin, additional, Li, Chunhao, additional, Li, Fen, additional, Li, Jun, additional, Li, Lanjuan, additional, Li, Ruofan, additional, Li, Ruoyu, additional, Li, Ting, additional, Li, Yang-Yang, additional, Li, Yanhua, additional, Li, Zhuorong, additional, Liang, Xiaomeng, additional, Liao, Guojian, additional, Lin, Ping, additional, Ling, Yun, additional, Liu, Bo, additional, Liu, Dongyou, additional, Liu, Guohua, additional, Liu, Huidi, additional, Liu, Jiafeng, additional, Liu, Jintao, additional, Liu, Qi, additional, Liu, Shu-Lin, additional, Liu, Taiping, additional, Liu, Tongbao, additional, Liu, Wei, additional, Liu, Yan, additional, Liu, Yanni, additional, Liu, Yisong, additional, Liu, Yuan, additional, Løbner-Olesen, Anders, additional, Loeffelholz, Michael, additional, Lu, Hongzhou, additional, Luna, Brian, additional, Ma, Bingting, additional, Ma, Chengying, additional, Ma, Shuang, additional, Ma, TianLi, additional, Madan, Rajat, additional, Mahle, Rachael E., additional, Mahlen, Steven D., additional, Malik, Satya Veer Singh, additional, Malik, Yashpal Singh, additional, Malvy, Denis, additional, Mann, Barbara J., additional, Marasini, Daya, additional, Maris, Alexander S., additional, Marjomäki, Varpu, additional, Marjuki, Henju, additional, Martín, Alberto, additional, Martín, Irene, additional, Martínez-Castillo, Moisés, additional, Martínez-Pabón, María Cecilia, additional, Mathison, Blaine A., additional, Ma’ayeh, Showgy, additional, McDowell, Andrew, additional, McLaughlin, Stephanie E., additional, McSheffrey, Gordon G., additional, Medrano, Francisco J., additional, Meehan, Conor J., additional, Mehta, Dhwani, additional, Mejía-Oquendo, Manuela, additional, Melo-Cristino, José, additional, Mendoza-Barberá, Elena, additional, Meng, Xinan, additional, Merino, Susana, additional, Merritt, Adam J., additional, Miller, Steve, additional, Miller, William R., additional, Minamino, Tohru, additional, Mirzaei, Mohammadali Khan, additional, Mora-Montes, Héctor M., additional, Mortensen, Joel, additional, Mostafa, Heba H., additional, Muhsen, Khitam, additional, Mujahed, Ahlam, additional, Muro, Antonio, additional, Murphy, Olwen C., additional, Newton, Hayley J., additional, Nguyen, April H., additional, Nichols, Wright W., additional, Niu, Siqiang, additional, Núñez, Félix, additional, Obregon, Dasiel, additional, Okamoto, Akira, additional, Okutani, Akiko, additional, Olabode, Abayomi, additional, Omar, Muna, additional, Ong, Edison, additional, Ouyang, Zhiming, additional, Pacak, Christina A., additional, Pacheco-Yépez, Judith, additional, Palmer, John, additional, Pang, Xiaoli, additional, Paredes-Sabja, Daniel, additional, Peng, Zhong, additional, Peng, Zonggen, additional, Pérez-Nevado, Francisco, additional, Poon, Art, additional, Pospíšilová, Petra, additional, Potts, Caelin C., additional, Pu, Qinqin, additional, Pujic, Petar, additional, Qi, Rui, additional, Qian, Chenyun, additional, Qian, Liu, additional, Qin, Aiping, additional, Qu, Fen, additional, Rakin, Alexander, additional, Ramesh, Ashwin, additional, Ramirez, Mario, additional, Rao, Yu, additional, Ratner, Adam J., additional, Rawool, Deepak B., additional, Rehman, Asma, additional, Ren, Jie, additional, Ren, Ping, additional, Retchless, Adam C., additional, Robertson, Erle S., additional, Rodríguez, Alicia, additional, Rodriguez, Azucena, additional, Rodríguez-Medina, Carolina, additional, Rodriguez-Nava, Veronica, additional, Rohde, Manfred, additional, Romero-Rodríguez, Alba, additional, Rosales-Morgan, Gabriela, additional, Rosenkranz, Andrea L., additional, Ruiz-Moyano, Santiago, additional, Ruokolainen, Visa, additional, Sabateen, Ali, additional, Sahu, Radhakrishna, additional, Sails, Andrew, additional, Sang, Yu, additional, Santana, Clarissa H., additional, Santos, Jesus A., additional, Santos, Renato L., additional, Schmitz, Jonathan E., additional, Serrano-Luna, Jesús, additional, Shen, Jianzhong, additional, Shen, Zhangqi, additional, Shibayama, Mineko, additional, Shirtliff, Mark E., additional, Silva-Costa, Catarina, additional, Silva-Olivares, Angélica, additional, Singh, Niraj Kumar, additional, Šmajs, David, additional, Smith, Robert P., additional, Smith, Sophie, additional, Snyder, Lori A.S., additional, Song, Yinggai, additional, Soro, Aurea Simon, additional, Spearman, Paul, additional, Spellberg, Brad, additional, Sprague, Lisa D., additional, Stratton, Charles W., additional, Strenk, Susan M., additional, Strugnell, Richard A., additional, Sun, Keer, additional, Suo, Xun, additional, Suzuki-Hatano, Silveli, additional, Svärd, Staffan, additional, Talbot, Elizabeth A., additional, Tamez-Castrellón, Alma K., additional, Tan, Nie, additional, Tang, Cynthia Y., additional, Tang, Yi-Wei, additional, Tao, Jia, additional, Tao, Lili, additional, Terrero-Salcedo, David, additional, Tharmalingam, Jayaraman, additional, Thwe, Phyu M., additional, Tiamani, Kawtar, additional, Tomás, Juan M., additional, Topaz, Nadav, additional, Tsai, Ang-Chen, additional, Tsalik, Ephraim L., additional, Tuomanen, Elaine I., additional, Turenne, Christine Y., additional, Tyagi, Anuj, additional, Uprety, Priyanka, additional, Valour, Florent, additional, van Hensbergen, Vincent P., additional, Venkatesan, Arun, additional, Vergis, Jess, additional, Villar, Margarita, additional, Vollmer, Waldemar, additional, Waites, Ken B., additional, Wan, Xiu-Feng, additional, Wang, Guiqing, additional, Wang, Lijun, additional, Wang, Lin, additional, Wang, Linqi, additional, Wang, Xiangru, additional, Wang, Xin, additional, Wang, Xinjie, additional, Wang, Ya-Ting, additional, Wang, Yang, additional, Wang, Yating, additional, Weil, Ana A., additional, Welkos, Susan L., additional, Wengenack, Nancy L., additional, Westblade, Lars F., additional, Whitfield, Chris, additional, Wu, Hui, additional, Wu, Lijuan, additional, Wu, Min, additional, Wu, Yarong, additional, Wu, Zhaowei, additional, Xiang, Ye, additional, Xiao, Di, additional, Xiao, Li, additional, Xiao, Lihua, additional, Xu, Tao, additional, Xu, Wenyue, additional, Xu, Xinping, additional, Xue, Jinling, additional, Yadav, Jay Prakash, additional, Yan, Junxiang, additional, Yan, Yixin, additional, Yang, Changmei, additional, Yang, Ruifu, additional, Yang, Ying, additional, Yao, Kaihu, additional, Yao, Yu-Feng, additional, Yeakle, Kyle C., additional, Yu, Demin, additional, Yu, Hao, additional, Yu, Xue-Jie, additional, Yuan, Zhenghong, additional, Zai, Wenjing, additional, Zhang, Jianzhong, additional, Zhang, Jing-Ren, additional, Zhang, Lanyue, additional, Zhang, Lijie, additional, Zhang, Qiwei, additional, Zhang, Wenbao, additional, Zhang, Wenhong, additional, Zhang, Xinxin, additional, Zhao, Youbao, additional, Zhou, Chuanmin, additional, Zhu, Feng, additional, Zhu, Jingting, additional, and Zhu, Yongqun, additional
- Published
- 2024
- Full Text
- View/download PDF
22. Improved DGCNN Based on Transformer for Point Cloud Segmentation
- Author
-
Zan, Guokuan, primary, Wang, Yang, additional, and Gao, Pengxiang, additional
- Published
- 2024
- Full Text
- View/download PDF
23. An Enhanced Downsampling Transformer Network for Point Cloud Semantic Segmentation
- Author
-
Wang, Yang, primary, Wei, Zixuan, additional, and Wan, Zhibo, additional
- Published
- 2024
- Full Text
- View/download PDF
24. Expectation asymmetries in mobile communication of Chinese "study mothers (Peidu Mama)"
- Author
-
Wang, Yang, primary
- Published
- 2023
- Full Text
- View/download PDF
25. Effect of biofilm formation on the Escherichia coli drug resistance of isolates from pigs in central China
- Author
-
Li, Jinpeng, Fan, Qingying, Mao, Chenlong, Jin, Manyu, Yi, Li, and Wang, Yang
- Published
- 2021
- Full Text
- View/download PDF
26. Harmonic LLMs are Trustworthy
- Author
-
Kersting, Nicholas S., Rahman, Mohammad, Vedala, Suchismitha, and Wang, Yang
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence ,Computer Science - Computation and Language ,Computer Science - Human-Computer Interaction - Abstract
We introduce an intuitive method to test the robustness (stability and explainability) of any black-box LLM in real-time, based upon the local deviation from harmoniticity, denoted as $\gamma$. To the best of our knowledge this is the first completely model-agnostic and unsupervised method of measuring the robustness of any given response from an LLM, based upon the model itself conforming to a purely mathematical standard. We conduct human annotation experiments to show the positive correlation of $\gamma$ with false or misleading answers, and demonstrate that following the gradient of $\gamma$ in stochastic gradient ascent efficiently exposes adversarial prompts. Measuring $\gamma$ across thousands of queries in popular LLMs (GPT-4, ChatGPT, Claude-2.1, Mixtral-8x7B, Smaug-72B, Llama2-7B, and MPT-7B) allows us to estimate the liklihood of wrong or hallucinatory answers automatically and quantitatively rank the reliability of these models in various objective domains (Web QA, TruthfulQA, and Programming QA). Across all models and domains tested, human ratings confirm that $\gamma \to 0$ indicates trustworthiness, and the low-$\gamma$ leaders among these models are GPT-4, ChatGPT, and Smaug-72B., Comment: 15 pages, 4 figures, 14 tables
- Published
- 2024
27. Real-Time 4K Super-Resolution of Compressed AVIF Images. AIS 2024 Challenge Survey
- Author
-
Conde, Marcos V., Lei, Zhijun, Li, Wen, Stejerean, Cosmin, Katsavounidis, Ioannis, Timofte, Radu, Yoon, Kihwan, Gankhuyag, Ganzorig, Lv, Jiangtao, Sun, Long, Pan, Jinshan, Dong, Jiangxin, Tang, Jinhui, Li, Zhiyuan, Wei, Hao, Ge, Chenyang, Zhang, Dongyang, Liu, Tianle, Chen, Huaian, Jin, Yi, Zhou, Menghan, Yan, Yiqiang, Gao, Si, Wu, Biao, Liu, Shaoli, Zheng, Chengjian, Zhang, Diankai, Wang, Ning, Qiu, Xintao, Zhou, Yuanbo, Wu, Kongxian, Dai, Xinwei, Tang, Hui, Deng, Wei, Gao, Qingquan, Tong, Tong, Lee, Jae-Hyeon, Choi, Ui-Jin, Yan, Min, Liu, Xin, Wang, Qian, Ye, Xiaoqian, Du, Zhan, Zhang, Tiansen, Peng, Long, Guo, Jiaming, Di, Xin, Liao, Bohao, Du, Zhibo, Xia, Peize, Pei, Renjing, Wang, Yang, Cao, Yang, Zha, Zhengjun, Han, Bingnan, Yu, Hongyuan, Wu, Zhuoyuan, Wan, Cheng, Liu, Yuqing, Yu, Haodong, Li, Jizhe, Huang, Zhijuan, Huang, Yuan, Zou, Yajun, Guan, Xianyu, Jia, Qi, Zhang, Heng, Yin, Xuanwu, Zuo, Kunlong, Moon, Hyeon-Cheol, Jeong, Tae-hyun, Yang, Yoonmo, Kim, Jae-Gon, Jeong, Jinwoo, and Kim, Sunjei
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Electrical Engineering and Systems Science - Image and Video Processing - Abstract
This paper introduces a novel benchmark as part of the AIS 2024 Real-Time Image Super-Resolution (RTSR) Challenge, which aims to upscale compressed images from 540p to 4K resolution (4x factor) in real-time on commercial GPUs. For this, we use a diverse test set containing a variety of 4K images ranging from digital art to gaming and photography. The images are compressed using the modern AVIF codec, instead of JPEG. All the proposed methods improve PSNR fidelity over Lanczos interpolation, and process images under 10ms. Out of the 160 participants, 25 teams submitted their code and models. The solutions present novel designs tailored for memory-efficiency and runtime on edge devices. This survey describes the best solutions for real-time SR of compressed high-resolution images., Comment: CVPR 2024, AI for Streaming (AIS) Workshop
- Published
- 2024
28. Delayed Bottlenecking: Alleviating Forgetting in Pre-trained Graph Neural Networks
- Author
-
Zhao, Zhe, Wang, Pengkun, Wang, Xu, Wen, Haibin, Xie, Xiaolong, Zhou, Zhengyang, Zhang, Qingfu, and Wang, Yang
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
Pre-training GNNs to extract transferable knowledge and apply it to downstream tasks has become the de facto standard of graph representation learning. Recent works focused on designing self-supervised pre-training tasks to extract useful and universal transferable knowledge from large-scale unlabeled data. However, they have to face an inevitable question: traditional pre-training strategies that aim at extracting useful information about pre-training tasks, may not extract all useful information about the downstream task. In this paper, we reexamine the pre-training process within traditional pre-training and fine-tuning frameworks from the perspective of Information Bottleneck (IB) and confirm that the forgetting phenomenon in pre-training phase may cause detrimental effects on downstream tasks. Therefore, we propose a novel \underline{D}elayed \underline{B}ottlenecking \underline{P}re-training (DBP) framework which maintains as much as possible mutual information between latent representations and training data during pre-training phase by suppressing the compression operation and delays the compression operation to fine-tuning phase to make sure the compression can be guided with labeled fine-tuning data and downstream tasks. To achieve this, we design two information control objectives that can be directly optimized and further integrate them into the actual model design. Extensive experiments on both chemistry and biology domains demonstrate the effectiveness of DBP.
- Published
- 2024
29. The Ninth NTIRE 2024 Efficient Super-Resolution Challenge Report
- Author
-
Ren, Bin, Li, Yawei, Mehta, Nancy, Timofte, Radu, Yu, Hongyuan, Wan, Cheng, Hong, Yuxin, Han, Bingnan, Wu, Zhuoyuan, Zou, Yajun, Liu, Yuqing, Li, Jizhe, He, Keji, Fan, Chao, Zhang, Heng, Zhang, Xiaolin, Yin, Xuanwu, Zuo, Kunlong, Liao, Bohao, Xia, Peizhe, Peng, Long, Du, Zhibo, Di, Xin, Li, Wangkai, Wang, Yang, Zhai, Wei, Pei, Renjing, Guo, Jiaming, Xu, Songcen, Cao, Yang, Zha, Zhengjun, Wang, Yan, Liu, Yi, Wang, Qing, Zhang, Gang, Zhang, Liou, Zhao, Shijie, Sun, Long, Pan, Jinshan, Dong, Jiangxin, Tang, Jinhui, Liu, Xin, Yan, Min, Wang, Qian, Zhou, Menghan, Yan, Yiqiang, Liu, Yixuan, Chan, Wensong, Tang, Dehua, Zhou, Dong, Wang, Li, Tian, Lu, Emad, Barsoum, Jia, Bohan, Qiao, Junbo, Zhou, Yunshuai, Zhang, Yun, Li, Wei, Lin, Shaohui, Zhou, Shenglong, Chen, Binbin, Liao, Jincheng, Zhao, Suiyi, Zhang, Zhao, Wang, Bo, Luo, Yan, Wei, Yanyan, Li, Feng, Wang, Mingshen, Guan, Jinhan, Hu, Dehua, Yu, Jiawei, Xu, Qisheng, Sun, Tao, Lan, Long, Xu, Kele, Lin, Xin, Yue, Jingtong, Yang, Lehan, Du, Shiyi, Qi, Lu, Ren, Chao, Han, Zeyu, Wang, Yuhan, Chen, Chaolin, Li, Haobo, Zheng, Mingjun, Yang, Zhongbao, Song, Lianhong, Yan, Xingzhuo, Fu, Minghan, Zhang, Jingyi, Li, Baiang, Zhu, Qi, Xu, Xiaogang, Guo, Dan, Guo, Chunle, Chen, Jiadi, Long, Huanhuan, Duanmu, Chunjiang, Lei, Xiaoyan, Liu, Jie, Jia, Weilin, Cao, Weifeng, Zhang, Wenlong, Mao, Yanyu, Guo, Ruilong, Zhang, Nihao, Pandey, Manoj, Chernozhukov, Maksym, Le, Giang, Cheng, Shuli, Wang, Hongyuan, Wei, Ziyan, Tang, Qingting, Wang, Liejun, Li, Yongming, Guo, Yanhui, Xu, Hao, Khatami-Rizi, Akram, Mahmoudi-Aznaveh, Ahmad, Hsu, Chih-Chung, Lee, Chia-Ming, Chou, Yi-Shiuan, Joshi, Amogh, Akalwadi, Nikhil, Malagi, Sampada, Yashaswini, Palani, Desai, Chaitra, Tabib, Ramesh Ashok, Patil, Ujwala, and Mudenagudi, Uma
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Electrical Engineering and Systems Science - Image and Video Processing - Abstract
This paper provides a comprehensive review of the NTIRE 2024 challenge, focusing on efficient single-image super-resolution (ESR) solutions and their outcomes. The task of this challenge is to super-resolve an input image with a magnification factor of x4 based on pairs of low and corresponding high-resolution images. The primary objective is to develop networks that optimize various aspects such as runtime, parameters, and FLOPs, while still maintaining a peak signal-to-noise ratio (PSNR) of approximately 26.90 dB on the DIV2K_LSDIR_valid dataset and 26.99 dB on the DIV2K_LSDIR_test dataset. In addition, this challenge has 4 tracks including the main track (overall performance), sub-track 1 (runtime), sub-track 2 (FLOPs), and sub-track 3 (parameters). In the main track, all three metrics (ie runtime, FLOPs, and parameter count) were considered. The ranking of the main track is calculated based on a weighted sum-up of the scores of all other sub-tracks. In sub-track 1, the practical runtime performance of the submissions was evaluated, and the corresponding score was used to determine the ranking. In sub-track 2, the number of FLOPs was considered. The score calculated based on the corresponding FLOPs was used to determine the ranking. In sub-track 3, the number of parameters was considered. The score calculated based on the corresponding parameters was used to determine the ranking. RLFN is set as the baseline for efficiency measurement. The challenge had 262 registered participants, and 34 teams made valid submissions. They gauge the state-of-the-art in efficient single-image super-resolution. To facilitate the reproducibility of the challenge and enable other researchers to build upon these findings, the code and the pre-trained model of validated solutions are made publicly available at https://github.com/Amazingren/NTIRE2024_ESR/., Comment: The report paper of NTIRE2024 Efficient Super-resolution, accepted by CVPRW2024
- Published
- 2024
30. Deep Learning for Cosmological Parameter Inference from Dark Matter Halo Density Field
- Author
-
Min, Zhiwei, Xiao, Xu, Ding, Jiacheng, Xiao, Liang, Jiang, Jie, Wu, Donglin, Lin, Qiufan, Li, Yin, Wang, Yang, Liu, Shuai, Chen, Zhixin, Li, Xiangru, Zhang, Jinqu, Zhang, Le, and Li, Xiao-Dong
- Subjects
Astrophysics - Cosmology and Nongalactic Astrophysics - Abstract
We propose a lightweight deep convolutional neural network (lCNN) to estimate cosmological parameters from simulated three-dimensional DM halo distributions and associated statistics. The training dataset comprises 2000 realizations of a cubic box with a side length of 1000 $h^{-1}{\rm Mpc}$, and interpolated over a cubic grid of $300^3$ voxels, with each simulation produced using $512^3$ DM particles and $512^3$ neutrinos . Under the flat $\Lambda$CDM model, simulations vary standard six cosmological parameters including $\Omega_m$, $\Omega_b$, $h$, $n_s$, $\sigma_8$, $w$, along with the neutrino mass sum, $M_\nu$. We find that: 1) within the framework of lCNN, extracting large-scale structure information is more efficient from the halo density field compared to relying on the statistical quantities including the power spectrum, the two-point correlation function, and the coefficients from wavelet scattering transform; 2) combining the halo density field with its Fourier transformed counterpart enhances predictions, while augmenting the training dataset with measured statistics further improves performance; 3) achieving high accuracy in inferring $\Omega_m$, $h$, $n_s$, and $\sigma_8$ by the neural network model, while being inefficient in predicting $\Omega_b$,$M_\nu$ and $w$; 4) compared to the simple random forest network trained with three statistical quantities, lCNN yields unbiased estimations with reduced statistical errors: approximately 33.3\% for $\Omega_m$, 20.0\% for $h$, 8.3\% for $n_s$, and 40.0\% for $\sigma_8$. Our study emphasizes this lCNN-based novel approach in extracting large-scale structure information and estimating cosmological parameters., Comment: 10 pages,9 figures
- Published
- 2024
31. IsoPredict: Dynamic Predictive Analysis for Detecting Unserializable Behaviors in Weakly Isolated Data Store Applications
- Author
-
Geng, Chujun, Blanas, Spyros, Bond, Michael D., and Wang, Yang
- Subjects
Computer Science - Programming Languages ,Computer Science - Databases - Abstract
This paper presents the first dynamic predictive analysis for data store applications under weak isolation levels, called Isopredict. Given an observed serializable execution of a data store application, Isopredict generates and solves SMT constraints to find an unserializable execution that is a feasible execution of the application. Isopredict introduces novel techniques that handle divergent application behavior; solve mutually recursive sets of constraints; and balance coverage, precision, and performance. An evaluation on four transactional data store benchmarks shows that Isopredict often predicts unserializable behaviors, 99% of which are feasible.
- Published
- 2024
- Full Text
- View/download PDF
32. Convergence Analysis of Flow Matching in Latent Space with Transformers
- Author
-
Jiao, Yuling, Lai, Yanming, Wang, Yang, and Yan, Bokai
- Subjects
Statistics - Machine Learning ,Computer Science - Machine Learning - Abstract
We present theoretical convergence guarantees for ODE-based generative models, specifically flow matching. We use a pre-trained autoencoder network to map high-dimensional original inputs to a low-dimensional latent space, where a transformer network is trained to predict the velocity field of the transformation from a standard normal distribution to the target latent distribution. Our error analysis demonstrates the effectiveness of this approach, showing that the distribution of samples generated via estimated ODE flow converges to the target distribution in the Wasserstein-2 distance under mild and practical assumptions. Furthermore, we show that arbitrary smooth functions can be effectively approximated by transformer networks with Lipschitz continuity, which may be of independent interest.
- Published
- 2024
33. Boosting Visual Recognition for Autonomous Driving in Real-world Degradations with Deep Channel Prior
- Author
-
Liu, Zhanwen, Li, Yuhang, Wang, Yang, Gao, Bolin, An, Yisheng, and Zhao, Xiangmo
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
The environmental perception of autonomous vehicles in normal conditions have achieved considerable success in the past decade. However, various unfavourable conditions such as fog, low-light, and motion blur will degrade image quality and pose tremendous threats to the safety of autonomous driving. That is, when applied to degraded images, state-of-the-art visual models often suffer performance decline due to the feature content loss and artifact interference caused by statistical and structural properties disruption of captured images. To address this problem, this work proposes a novel Deep Channel Prior (DCP) for degraded visual recognition. Specifically, we observe that, in the deep representation space of pre-trained models, the channel correlations of degraded features with the same degradation type have uniform distribution even if they have different content and semantics, which can facilitate the mapping relationship learning between degraded and clear representations in high-sparsity feature space. Based on this, a novel plug-and-play Unsupervised Feature Enhancement Module (UFEM) is proposed to achieve unsupervised feature correction, where the multi-adversarial mechanism is introduced in the first stage of UFEM to achieve the latent content restoration and artifact removal in high-sparsity feature space. Then, the generated features are transferred to the second stage for global correlation modulation under the guidance of DCP to obtain high-quality and recognition-friendly features. Evaluations of three tasks and eight benchmark datasets demonstrate that our proposed method can comprehensively improve the performance of pre-trained models in real degradation conditions. The source code is available at https://github.com/liyuhang166/Deep_Channel_Prior
- Published
- 2024
34. Meta Episodic learning with Dynamic Task Sampling for CLIP-based Point Cloud Classification
- Author
-
Ghose, Shuvozit and Wang, Yang
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Point cloud classification refers to the process of assigning semantic labels or categories to individual points within a point cloud data structure. Recent works have explored the extension of pre-trained CLIP to 3D recognition. In this direction, CLIP-based point cloud models like PointCLIP, CLIP2Point have become state-of-the-art methods in the few-shot setup. Although these methods show promising performance for some classes like airplanes, desks, guitars, etc, the performance for some classes like the cup, flower pot, sink, nightstand, etc is still far from satisfactory. This is due to the fact that the adapter of CLIP-based models is trained using randomly sampled N-way K-shot data in the standard supervised learning setup. In this paper, we propose a novel meta-episodic learning framework for CLIP-based point cloud classification, addressing the challenges of limited training examples and sampling unknown classes. Additionally, we introduce dynamic task sampling within the episode based on performance memory. This sampling strategy effectively addresses the challenge of sampling unknown classes, ensuring that the model learns from a diverse range of classes and promotes the exploration of underrepresented categories. By dynamically updating the performance memory, we adaptively prioritize the sampling of classes based on their performance, enhancing the model's ability to handle challenging and real-world scenarios. Experiments show an average performance gain of 3-6\% on ModelNet40 and ScanobjectNN datasets in a few-shot setup.
- Published
- 2024
35. Structure Matters: Tackling the Semantic Discrepancy in Diffusion Models for Image Inpainting
- Author
-
Liu, Haipeng, Wang, Yang, Qian, Biao, Wang, Meng, and Rui, Yong
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Denoising diffusion probabilistic models for image inpainting aim to add the noise to the texture of image during the forward process and recover masked regions with unmasked ones of the texture via the reverse denoising process. Despite the meaningful semantics generation, the existing arts suffer from the semantic discrepancy between masked and unmasked regions, since the semantically dense unmasked texture fails to be completely degraded while the masked regions turn to the pure noise in diffusion process, leading to the large discrepancy between them. In this paper, we aim to answer how unmasked semantics guide texture denoising process;together with how to tackle the semantic discrepancy, to facilitate the consistent and meaningful semantics generation. To this end, we propose a novel structure-guided diffusion model named StrDiffusion, to reformulate the conventional texture denoising process under structure guidance to derive a simplified denoising objective for image inpainting, while revealing: 1) the semantically sparse structure is beneficial to tackle semantic discrepancy in early stage, while dense texture generates reasonable semantics in late stage; 2) the semantics from unmasked regions essentially offer the time-dependent structure guidance for the texture denoising process, benefiting from the time-dependent sparsity of the structure semantics. For the denoising process, a structure-guided neural network is trained to estimate the simplified denoising objective by exploiting the consistency of the denoised structure between masked and unmasked regions. Besides, we devise an adaptive resampling strategy as a formal criterion as whether structure is competent to guide the texture denoising process, while regulate their semantic correlations. Extensive experiments validate the merits of StrDiffusion over the state-of-the-arts. Our code is available at https://github.com/htyjers/StrDiffusion., Comment: 15 pages, 10 figures, to appear CVPR 2024
- Published
- 2024
36. Toward CXL-Native Memory Tiering via Device-Side Profiling
- Author
-
Zhou, Zhe, Chen, Yiqi, Zhang, Tao, Wang, Yang, Shu, Ran, Xu, Shuotao, Cheng, Peng, Qu, Lei, Xiong, Yongqiang, and Sun, Guangyu
- Subjects
Computer Science - Hardware Architecture - Abstract
The Compute Express Link (CXL) interconnect has provided the ability to integrate diverse memory types into servers via byte-addressable SerDes links. Harnessing the full potential of such heterogeneous memory systems requires efficient memory tiering. However, existing research in this domain has been constrained by low-resolution and high-overhead memory access profiling techniques. To address this critical challenge, we propose to enhance existing memory tiering systems with a novel NeoMem solution. NeoMem offloads memory profiling functions to device-side controllers, integrating a dedicated hardware unit called NeoProf. NeoProf readily tracks memory access and provides the operating system with crucial page hotness statistics and other useful system state information. On the OS kernel side, we introduce a revamped memory-tiering strategy, enabling accurate and timely hot page promotion based on NeoProf statistics. We implement NeoMem on a real CXL-enabled FPGA platform and Linux kernel v6.3. Comprehensive evaluations demonstrate that NeoMem achieves 32% to 67% geomean speedup over several existing memory tiering solutions.
- Published
- 2024
37. Lightweight Embeddings for Graph Collaborative Filtering
- Author
-
Liang, Xurong, Chen, Tong, Cui, Lizhen, Wang, Yang, Wang, Meng, and Yin, Hongzhi
- Subjects
Computer Science - Information Retrieval - Abstract
Graph neural networks (GNNs) are currently one of the most performant collaborative filtering methods. Meanwhile, owing to the use of an embedding table to represent each user/item as a distinct vector, GNN-based recommenders have inherited the long-standing defect of parameter inefficiency. As a common practice for scalable embeddings, parameter sharing enables the use of fewer embedding vectors (i.e., meta-embeddings). When assigning meta-embeddings, most existing methods are a heuristically designed, predefined mapping from each user's/item's ID to the corresponding meta-embedding indexes, thus simplifying the optimization problem into learning only the meta-embeddings. However, in the context of GNN-based collaborative filtering, such a fixed mapping omits the semantic correlations between entities that are evident in the user-item interaction graph, leading to suboptimal recommendation performance. To this end, we propose Lightweight Embeddings for Graph Collaborative Filtering (LEGCF), a parameter-efficient embedding framework dedicated to GNN-based recommenders. LEGCF innovatively introduces an assignment matrix as an extra learnable component on top of meta-embeddings. To jointly optimize these two heavily entangled components, aside from learning the meta-embeddings by minimizing the recommendation loss, LEGCF further performs efficient assignment update by enforcing a novel semantic similarity constraint and finding its closed-form solution based on matrix pseudo-inverse. The meta-embeddings and assignment matrix are alternately updated, where the latter is sparsified on the fly to ensure negligible storage overhead. Extensive experiments on three benchmark datasets have verified LEGCF's smallest trade-off between size and performance, with consistent accuracy gain over state-of-the-art baselines. The codebase of LEGCF is available in https://github.com/xurong-liang/LEGCF., Comment: Accepted by SIGIR '24
- Published
- 2024
38. Remote Cooling of Spin-ensembles through a Spin-mechanical Hybrid Interface
- Author
-
Wang, Yang, Dasari, Durga Bhaktavatsala Rao, and Wrachtrup, Jörg
- Subjects
Quantum Physics - Abstract
We present a protocol for the ground-state cooling of a tripartite hybrid quantum system, in which a macroscopic oscillator acts as a mediator between a single probe spin and a remote spin ensemble. In the presence of weak dispersive coupling between the spins and the oscillator, cooling of the oscillator and the ensemble spins can be achieved by exploiting the feedback from frequent measurements of the single probe spin. We explore the parameter regimes necessary to cool the ensemble, the oscillator, or both to their thermal ground states. This novel cooling protocol shows that, even with only weak dispersive coupling, energy transfer-like effects can be obtained by simply manipulating the probe spin. These results not only contribute to the development of a practical solution for cooling/polarizing large spin ensembles, but also provide a relatively simple means of tuning the dynamics of a hybrid system. The proposed cooling protocol thus has broader implications for advancing various quantum technology applications, such as macroscopic quantum state generation and remote sensing.
- Published
- 2024
39. Anomalous thermal conductivity in 2D silica nanocages of immobilizing noble gas atom
- Author
-
Wang, Yang, Gao, Zhibin, Wang, Xiaoying, Sun, Jinping, Feng, Minxuan, Hao, Yuzhou, Li, Xuejie, Zhao, Yinchang, and Ding, Xiangdong
- Subjects
Condensed Matter - Materials Science ,Physics - Computational Physics - Abstract
Noble gas atoms such as Kr and Xe are byproducts of nuclear fission in nuclear plants. How to trap and confine these volatile even radioactive gases is particularly challenging. Recent studies have shown that they can be trapped in nanocages of ultrathin silica. Here, we exhibit with self-consistent phonon theory and four-phonon (4ph) scattering where the adsorption of noble gases results in an anomalous increase in lattice thermal conductivity, while the presence of Cu atoms doping leads to a reduction in lattice thermal conductivity. We trace this behavior in host-guest 2D silica to an interplay of tensile strain, rattling phonon modes, and redistribution of electrons. We also find that 4ph scatterings play indispensable roles in the lattice thermal conductivity of 2D silica. Our work illustrates the microscopic heat transfer mechanism in 2D silica nanocages with the immobilization of noble gas atoms and inspires further exploring materials with the kagome and glasslike lattice thermal conductivity., Comment: 7 pages, 4 figures
- Published
- 2024
- Full Text
- View/download PDF
40. Spatio-Temporal Fluid Dynamics Modeling via Physical-Awareness and Parameter Diffusion Guidance
- Author
-
Wu, Hao, Xu, Fan, Duan, Yifan, Niu, Ziwei, Wang, Weiyan, Lu, Gaofeng, Wang, Kun, Liang, Yuxuan, and Wang, Yang
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence ,Physics - Fluid Dynamics - Abstract
This paper proposes a two-stage framework named ST-PAD for spatio-temporal fluid dynamics modeling in the field of earth sciences, aiming to achieve high-precision simulation and prediction of fluid dynamics through spatio-temporal physics awareness and parameter diffusion guidance. In the upstream stage, we design a vector quantization reconstruction module with temporal evolution characteristics, ensuring balanced and resilient parameter distribution by introducing general physical constraints. In the downstream stage, a diffusion probability network involving parameters is utilized to generate high-quality future states of fluids, while enhancing the model's generalization ability by perceiving parameters in various physical setups. Extensive experiments on multiple benchmark datasets have verified the effectiveness and robustness of the ST-PAD framework, which showcase that ST-PAD outperforms current mainstream models in fluid dynamics modeling and prediction, especially in effectively capturing local representations and maintaining significant advantages in OOD generations.
- Published
- 2024
41. RL in Markov Games with Independent Function Approximation: Improved Sample Complexity Bound under the Local Access Model
- Author
-
Fan, Junyi, Han, Yuxuan, Zeng, Jialin, Cai, Jian-Feng, Wang, Yang, Xiang, Yang, and Zhang, Jiheng
- Subjects
Computer Science - Machine Learning - Abstract
Efficiently learning equilibria with large state and action spaces in general-sum Markov games while overcoming the curse of multi-agency is a challenging problem. Recent works have attempted to solve this problem by employing independent linear function classes to approximate the marginal $Q$-value for each agent. However, existing sample complexity bounds under such a framework have a suboptimal dependency on the desired accuracy $\varepsilon$ or the action space. In this work, we introduce a new algorithm, Lin-Confident-FTRL, for learning coarse correlated equilibria (CCE) with local access to the simulator, i.e., one can interact with the underlying environment on the visited states. Up to a logarithmic dependence on the size of the state space, Lin-Confident-FTRL learns $\epsilon$-CCE with a provable optimal accuracy bound $O(\epsilon^{-2})$ and gets rids of the linear dependency on the action space, while scaling polynomially with relevant problem parameters (such as the number of agents and time horizon). Moreover, our analysis of Linear-Confident-FTRL generalizes the virtual policy iteration technique in the single-agent local planning literature, which yields a new computationally efficient algorithm with a tighter sample complexity bound when assuming random access to the simulator., Comment: Accepted at the 27th International Conference on Artificial Intelligence and Statistics (AISTATS 2024)
- Published
- 2024
42. FairSTG: Countering performance heterogeneity via collaborative sample-level optimization
- Author
-
Lin, Gengyu, Zhou, Zhengyang, Huang, Qihe, Yang, Kuo, Cheng, Shifen, and Wang, Yang
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
Spatiotemporal learning plays a crucial role in mobile computing techniques to empower smart cites. While existing research has made great efforts to achieve accurate predictions on the overall dataset, they still neglect the significant performance heterogeneity across samples. In this work, we designate the performance heterogeneity as the reason for unfair spatiotemporal learning, which not only degrades the practical functions of models, but also brings serious potential risks to real-world urban applications. To fix this gap, we propose a model-independent Fairness-aware framework for SpatioTemporal Graph learning (FairSTG), which inherits the idea of exploiting advantages of well-learned samples to challenging ones with collaborative mix-up. Specifically, FairSTG consists of a spatiotemporal feature extractor for model initialization, a collaborative representation enhancement for knowledge transfer between well-learned samples and challenging ones, and fairness objectives for immediately suppressing sample-level performance heterogeneity. Experiments on four spatiotemporal datasets demonstrate that our FairSTG significantly improves the fairness quality while maintaining comparable forecasting accuracy. Case studies show FairSTG can counter both spatial and temporal performance heterogeneity by our sample-level retrieval and compensation, and our work can potentially alleviate the risks on spatiotemporal resource allocation for underrepresented urban regions., Comment: Under review by IEEE Transactions on Mobile Computing
- Published
- 2024
43. LAN: Learning Adaptive Neighbors for Real-Time Insider Threat Detection
- Author
-
Cai, Xiangrui, Wang, Yang, Xu, Sihan, Li, Hao, Zhang, Ying, Liu, Zheli, and Yuan, Xiaojie
- Subjects
Computer Science - Cryptography and Security ,Computer Science - Artificial Intelligence ,Computer Science - Machine Learning - Abstract
Enterprises and organizations are faced with potential threats from insider employees that may lead to serious consequences. Previous studies on insider threat detection (ITD) mainly focus on detecting abnormal users or abnormal time periods (e.g., a week or a day). However, a user may have hundreds of thousands of activities in the log, and even within a day there may exist thousands of activities for a user, requiring a high investigation budget to verify abnormal users or activities given the detection results. On the other hand, existing works are mainly post-hoc methods rather than real-time detection, which can not report insider threats in time before they cause loss. In this paper, we conduct the first study towards real-time ITD at activity level, and present a fine-grained and efficient framework LAN. Specifically, LAN simultaneously learns the temporal dependencies within an activity sequence and the relationships between activities across sequences with graph structure learning. Moreover, to mitigate the data imbalance problem in ITD, we propose a novel hybrid prediction loss, which integrates self-supervision signals from normal activities and supervision signals from abnormal activities into a unified loss for anomaly detection. We evaluate the performance of LAN on two widely used datasets, i.e., CERT r4.2 and CERT r5.2. Extensive and comparative experiments demonstrate the superiority of LAN, outperforming 9 state-of-the-art baselines by at least 9.92% and 6.35% in AUC for real-time ITD on CERT r4.2 and r5.2, respectively. Moreover, LAN can be also applied to post-hoc ITD, surpassing 8 competitive baselines by at least 7.70% and 4.03% in AUC on two datasets. Finally, the ablation study, parameter analysis, and compatibility analysis evaluate the impact of each module and hyper-parameter in LAN. The source code can be obtained from https://github.com/Li1Neo/LAN., Comment: 13 pages
- Published
- 2024
44. RulePrompt: Weakly Supervised Text Classification with Prompting PLMs and Self-Iterative Logical Rules
- Author
-
Li, Miaomiao, Zhu, Jiaqi, Wang, Yang, Yang, Yi, Li, Yilin, and Wang, Hongan
- Subjects
Computer Science - Computation and Language - Abstract
Weakly supervised text classification (WSTC), also called zero-shot or dataless text classification, has attracted increasing attention due to its applicability in classifying a mass of texts within the dynamic and open Web environment, since it requires only a limited set of seed words (label names) for each category instead of labeled data. With the help of recently popular prompting Pre-trained Language Models (PLMs), many studies leveraged manually crafted and/or automatically identified verbalizers to estimate the likelihood of categories, but they failed to differentiate the effects of these category-indicative words, let alone capture their correlations and realize adaptive adjustments according to the unlabeled corpus. In this paper, in order to let the PLM effectively understand each category, we at first propose a novel form of rule-based knowledge using logical expressions to characterize the meanings of categories. Then, we develop a prompting PLM-based approach named RulePrompt for the WSTC task, consisting of a rule mining module and a rule-enhanced pseudo label generation module, plus a self-supervised fine-tuning module to make the PLM align with this task. Within this framework, the inaccurate pseudo labels assigned to texts and the imprecise logical rules associated with categories mutually enhance each other in an alternative manner. That establishes a self-iterative closed loop of knowledge (rule) acquisition and utilization, with seed words serving as the starting point. Extensive experiments validate the effectiveness and robustness of our approach, which markedly outperforms state-of-the-art weakly supervised methods. What is more, our approach yields interpretable category rules, proving its advantage in disambiguating easily-confused categories., Comment: Accepted by WWW 2024
- Published
- 2024
45. ComS2T: A complementary spatiotemporal learning system for data-adaptive model evolution
- Author
-
Zhou, Zhengyang, Huang, Qihe, Wang, Binwu, Hou, Jianpeng, Yang, Kuo, Liang, Yuxuan, and Wang, Yang
- Subjects
Computer Science - Machine Learning - Abstract
Spatiotemporal (ST) learning has become a crucial technique to enable smart cities and sustainable urban development. Current ST learning models capture the heterogeneity via various spatial convolution and temporal evolution blocks. However, rapid urbanization leads to fluctuating distributions in urban data and city structures over short periods, resulting in existing methods suffering generalization and data adaptation issues. Despite efforts, existing methods fail to deal with newly arrived observations and those methods with generalization capacity are limited in repeated training. Motivated by complementary learning in neuroscience, we introduce a prompt-based complementary spatiotemporal learning termed ComS2T, to empower the evolution of models for data adaptation. ComS2T partitions the neural architecture into a stable neocortex for consolidating historical memory and a dynamic hippocampus for new knowledge update. We first disentangle two disjoint structures into stable and dynamic weights, and then train spatial and temporal prompts by characterizing distribution of main observations to enable prompts adaptive to new data. This data-adaptive prompt mechanism, combined with a two-stage training process, facilitates fine-tuning of the neural architecture conditioned on prompts, thereby enabling efficient adaptation during testing. Extensive experiments validate the efficacy of ComS2T in adapting to various spatiotemporal out-of-distribution scenarios while maintaining efficient inference capabilities.
- Published
- 2024
46. Mirror real Chern insulator in two and three dimensions
- Author
-
Wang, Yang, Cui, Chaoxi, Zhang, Run-Wu, Wang, Xiaotian, Yu, Zhi-Ming, Liu, Gui-Bin, and Yao, Yugui
- Subjects
Condensed Matter - Materials Science - Abstract
A real Chern insulator (RCI) featuring a real Chern number and a second-order boundary mode appears in a two-dimensional (2D) system with the space-time inversion symmetry (PT ). Here, we propose a kind of RCI: mirror real Chern insulator (MRCI) which emerges from the system having additional horizontal mirror symmetry Mz. The MRCI generally is characterized by two independent real Chern numbers, respectively defined in the two mirror subsystems of the system. Hence, the MRCI may host the second-order boundary modes different from the conventional RCI. We show that for spinless systems, the definition of the MRCI is straightforward, as PT keeps each mirror subsystem invariant. For the spinful systems with both PT and Mz, the real Chern number for the total system remain well defined, as MzPT = C2zT , and (C2zT )2= 1. However, since C2zT exchanges the two mirror subsystems, the definition of the MRCI in spinful systems requires the help of projective symmetry algebra. We also discuss the MRCIs in 3D systems, where the MRCI is defined on certain mirror-invariant 2D planes. Compared with its 2D counterpart, the 3D MRCI can exhibit more abundant physics when the systems have additional nonsymmorphic operators. Several concrete MRCI models including 2D and 3D, spinless and spinful models are constructed to further demonstrate our ideas.
- Published
- 2024
47. Graph Pointer Network and Reinforcement Learning for Thinnest Path Problem
- Author
-
Li, Jingjing, primary, Wang, Yang, additional, and Zhang, Chuang, additional
- Published
- 2023
- Full Text
- View/download PDF
48. A glimpse of the multiple applications of virtual reality
- Author
-
Wang, Yang, additional, Massalee, Rachel, additional, Beebe, Kris, additional, Latham, William, additional, Putnam, Lance, additional, Westendorp, Janine, additional, Evers, Andrea, additional, Testa, Marco, additional, Silber, Denise, additional, Trost, Zina, additional, Colloca, Giancarlo, additional, and Colloca, Luana, additional
- Published
- 2023
- Full Text
- View/download PDF
49. Representations of Socialist Mobility in Post-WWII China-Italy Cultural Exchange
- Author
-
Wang, Yang, primary and Tanga, Martina, additional
- Published
- 2023
- Full Text
- View/download PDF
50. Study on the Relief Response of Cross-Regional Natural Disaster Risk
- Author
-
Wang, Yang, primary, Ning, Baokun, additional, Chen, Zhuo, additional, Zhang, Shuai, additional, and Deng, Duo, additional
- Published
- 2023
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.