Search

This paper presents a comprehensive study and improvement of the Restormer architecture for high-resolution image motion deblurring. We introduce architectural modifications that reduce model complexity by 18.4% while maintaining or improving performance through optimized attention mechanisms. Our enhanced training pipeline incorporates additional transformations including color jitter, Gaussian blur, and perspective transforms to improve model robustness as well as a new frequency loss term. Extensive experiments on the RealBlur-R, RealBlur-J, and Ultra-High-Definition Motion blurred (UHDM) datasets demonstrate the effectiveness of our approach. The improved architecture shows better convergence behavior and reduced training time while maintaining competitive performance across challenging scenarios. We also provide detailed ablation studies analyzing the impact of our modifications on model behavior and performance. Our results suggest that thoughtful architectural simplification combined with enhanced training strategies can yield more efficient yet equally capable models for motion deblurring tasks. Code and Data Available at: https://github.com/hamzafer/image-deblurring, Comment: 14 pages, 18 figures Submitted as a preprint, no prior journal/conference submission

The development of color centers in silicon enables scalable quantum technologies by combining telecom-wavelength emission and compatibility with mature silicon fabrication. However, large-scale integration requires precise control of each emitter's optical transition to generate indistinguishable photons for quantum networking. Here, we demonstrate a foundry-fabricated photonic integrated circuit (PIC) combining suspended silicon waveguides with a microelectromechanical (MEMS) cantilever to apply local strain and spectrally tune individual G-centers. Applying up to 35 V between the cantilever and the substrate induces a reversible wavelength shift of the zero-phonon line exceeding 100 pm, with no loss in brightness. Moreover, by modeling the strain-induced shifts with a `digital twin' physical model, we achieve vertical localization of color centers with sub-3 nm vertical resolution, directly correlating their spatial position, dipole orientation, and spectral behavior. This method enables on-demand, low-power control of emission spectrum and nanoscale localization of color centers, advancing quantum networks on a foundry-compatible platform., Comment: 30 pages, 28 figures

In this paper, we calculate the unit groups and the $2$-class numbers of the fields $ \mathbb{K}= \mathbb{Q}(\sqrt{2}, \sqrt{p_1}, \sqrt{p_2})$ and $ \mathbb{L}= \mathbb{Q}( \sqrt{-1},\sqrt{2}, \sqrt{p_1}, \sqrt{p_2})$, where $p_1$ and $p_2$ are two prime numbers satisfying $p_1\equiv p_2 \equiv 1 \pmod{4}$., Comment: 36 pages

The introduction of new features and services in the banking sector often overwhelms customers, creating an opportunity for banks to enhance user experience through financial chatbots powered by large language models (LLMs). We initiated an AI agent designed to provide customers with relevant information about banking services and insights from annual reports. We proposed a hybrid Customer Analysis Pipeline Retrieval-Augmented Generation (CAPRAG) that effectively addresses both relationship-based and contextual queries, thereby improving customer engagement in the digital banking landscape. To implement this, we developed a processing pipeline to refine text data, which we utilized in two main frameworks: Vector RAG and Graph RAG. This dual approach enables us to populate both vector and graph databases with processed data for efficient retrieval. The Cypher query component is employed to effectively query the graph database. When a user submits a query, it is first expanded by a query expansion module before being routed to construct a final query from the hybrid Knowledge Base (KB). This final query is then sent to an open-source LLM for response generation. Overall, our innovative, designed to international banks, serves bank's customers in an increasingly complex digital environment, enhancing clarity and accessibility of information., Comment: 14 pages, 5 Figures, 3 Tables

Phase modulation is demonstrated in a quantum Stark effect modulator designed to operate in the mid-infrared at wavelength around 10 $\mu$m. Both phase and amplitude modulation are simultaneously resolved through the measurement of the heterodyne signal arising from the beating of a quantum cascade laser with a highly stabilized frequency comb. The highest measured phase shift is more than 5 degrees with an associated intensity modulation of 5%. The experimental results are in full agreement with our model in which the complex susceptibility is precisely described considering the linear voltage dependent Stark shift of the optical resonance., Comment: 8 pages, 5 figures

The rapid expansion of the industrial Internet of things (IIoT) has introduced new challenges in securing critical infrastructures against sophisticated cyberthreats. This study presents the development and evaluation of an advanced Intrusion detection (IDS) based on a hybrid LSTM-convolution neural network (CNN)-Attention architecture, specifically designed to detect and classify cyberattacks in IIoT environments. The research focuses on two key classification tasks: binary and multi-class classification. The proposed models was rigorously tested using the Edge-IIoTset dataset. To mitigate the class imbalance in the dataset, the synthetic minority over-sampling technique (SMOTE) was employed to generate synthetic samples for the underrepresented classes. This ensured that the model could learn effectively from all classes, thereby improving the overall classification performance. Through systematic experimentation, various deep learning (DL) models were compared, ultimately demonstrating that the LSTM-CNN-Attention model consistently outperformed others across key performance metrics. In binary classification, the model achieved near-perfect accuracy, while in multi-class classification, it maintained a high accuracy level (99.04%), effectively categorizing different attack types with a loss value of 0.0220%.

In this paper, we tackle the challenge of instance segmentation for foreign objects in chest radiographs, commonly seen in postoperative follow-ups with stents, pacemakers, or ingested objects in children. The diversity of foreign objects complicates dense annotation, as shown in insufficient existing datasets. To address this, we propose the simple generation of synthetic data through (1) insertion of arbitrary shapes (lines, polygons, ellipses) with varying contrasts and opacities, and (2) cut-paste augmentations from a small set of semi-automatically extracted labels. These insertions are guided by anatomy labels to ensure realistic placements, such as stents appearing only in relevant vessels. Our approach enables networks to segment complex structures with minimal manually labeled data. Notably, it achieves performance comparable to fully supervised models while using 93\% fewer manual annotations.

Magnetoencephalography (MEG) is a cutting-edge neuroimaging technique that measures the intricate brain dynamics underlying cognitive processes with an unparalleled combination of high temporal and spatial precision. MEG data analytics has always relied on advanced signal processing and mathematical and statistical tools for various tasks ranging from data cleaning to probing the signals' rich dynamics and estimating the neural sources underlying the surface-level recordings. Like in most domains, the surge in Artificial Intelligence (AI) has led to the increased use of Machine Learning (ML) methods for MEG data classification. More recently, an emerging trend in this field is using Artificial Neural Networks (ANNs) to address many MEG-related tasks. This review provides a comprehensive overview of how ANNs are being used with MEG data from three vantage points: First, we review work that employs ANNs for MEG signal classification, i.e., for brain decoding. Second, we report on work that has used ANNs as putative models of information processing in the human brain. Finally, we examine studies that use ANNs as techniques to tackle methodological questions in MEG, including artifact correction and source estimation. Furthermore, we assess the current strengths and limitations of using ANNs with MEG and discuss future challenges and opportunities in this field. Finally, by establishing a detailed portrait of the field and providing practical recommendations for the future, this review seeks to provide a helpful reference for both seasoned MEG researchers and newcomers to the field who are interested in using ANNs to enhance the exploration of the complex dynamics of the human brain with MEG., Comment: Full table containing information of papers reviewed can be found in this google sheet: https://tinyurl.com/ub3s5mr

The Queue-Reactive model introduced by Huang et al. (2015) has become a standard tool for limit order book modeling, widely adopted by both researchers and practitioners for its simplicity and effectiveness. We present the Multidimensional Deep Queue-Reactive (MDQR) model, which extends this framework in three ways: it relaxes the assumption of queue independence, enriches the state space with market features, and models the distribution of order sizes. Through a neural network architecture, the model learns complex dependencies between different price levels and adapts to varying market conditions, while preserving the interpretable point-process foundation of the original framework. Using data from the Bund futures market, we show that MDQR captures key market properties including the square-root law of market impact, cross-queue correlations, and realistic order size patterns. The model demonstrates particular strength in reproducing both conditional and stationary distributions of order sizes, as well as various stylized facts of market microstructure. The model achieves this while maintaining the computational efficiency needed for practical applications such as strategy development through reinforcement learning or realistic backtesting.

Under-frequency load shedding (UFLS) prevents system collapse during large disturbances. Increased penetration of distributed energy resources (DERs) and reduced system inertia makes it challenging to design a static UFLS scheme, which relies on preset frequency thresholds and load shed fractions to meet design criteria across all possible operating conditions. Due to non-linearity and traceability issues, previous adaptive UFLS schemes use simplified tractable frequency models that overlook AC network effects such as voltage-dependent load/generation. This paper leverages model order reduction techniques to obtain a higher fidelity low-order model of system frequency dynamics that captures AC network effects while incorporating turbine governor action and their associated limits. The model is then used in a new AC-aware predictive optimization framework to adapt UFLS setpoints periodically based on current operating conditions while minimizing load shed. Validated on a 1,648-bus system with PSS/E simulations, the proposed method meets design criteria under various operating conditions and disturbance scenarios. Furthermore, the framework outperforms conventional static UFLS schemes and adaptive UFLS schemes based on simplified dynamic models.

Recent demonstrations indicate that silicon-spin QPUs will be able to shuttle physical qubits rapidly and with high fidelity - a desirable feature for maximising logical connectivity, supporting new codes, and routing around damage. However it may seem that shuttling at the logical level is unwise: static defects in the device may 'scratch' a logical qubit as it passes, causing correlated errors to which the code is highly vulnerable. Here we explore an architecture where logical qubits are 1D strings ('snakes') which can be moved freely over a planar latticework. Possible scratch events are inferred via monitor qubits and the complimentary gap; if deemed a risk, remarkably the shuttle process can be undone in a way that negates any corruption. Interaction between logical snakes is facilitated by a semi-transversal method. We obtain encouraging estimates for the tolerable levels of shuttling-related imperfections., Comment: 27 pages, 23 figures

A (molecular) graph in which all vertices have the same degree is known as a regular graph. According to Gutman, Hansen, and M\'elot [J. Chem. Inf. Model. 45 (2005) 222-230], it is of interest to measure the irregularity of nonregular molecular graphs both for descriptive purposes and for QSAR/QSPR studies. The graph invariants that can be used to measure the irregularity of graphs are referred to as irregularity measures. One of the well-studied irregularity measures is the ``total irregularity'' measure, which was introduced about a decade ago. Bounds and optimization problems for this measure have already been extensively studied. A considerable number of existing results (concerning this measure) also hold for molecular graphs; particularly, the ones regarding lower bounds and minimum values of the mentioned measure. The primary objective of the present review article is to collect the existing bounds and optimal results concerning the total irregularity measure. Several open problems related to the existing results on the total irregularity measure are also given., Comment: 25 pages, 2 figures

In this paper, we analyze examples of research institutes that stand out in scientific excellence and social impact. We define key practices for evaluating research results, economic conditions, and the selection of specific research topics. Special focus is placed on small countries and the field of artificial intelligence. The aim is to identify components that enable institutes to achieve a high level of innovation, self-sustainability, and social benefits., Comment: 20 pages, 1 table, Bosnian language. Presented at Naucno-strucna konferencija o umjetnoj inteligenciji. Federalno ministarstvo obrazovanja i nauke, Mostar, Bosna i Hercegovina, May 2024

Informed by the success of the transformer model in various computer vision tasks, we design an end-to-end trainable model for the automatic detection and classification of bleeding and non-bleeding frames extracted from Wireless Capsule Endoscopy (WCE) videos. Based on the DETR model, our model uses the Resnet50 for feature extraction, the transformer encoder-decoder for bleeding and non-bleeding region detection, and a feedforward neural network for classification. Trained in an end-to-end approach on the Auto-WCEBleedGen Version 1 challenge training set, our model performs both detection and classification tasks as a single unit. Our model achieves an accuracy, recall, and F1-score classification percentage score of 98.28, 96.79, and 98.37 respectively, on the Auto-WCEBleedGen version 1 validation set. Further, we record an average precision (AP @ 0.5), mean-average precision (mAP) of 0.7447 and 0.7328 detection results. This earned us a 3rd place position in the challenge. Our code is publicly available via https://github.com/BasitAlawode/WCEBleedGen.

Peer-to-peer (P2P) energy trading, Smart Grids (SG), and electric vehicle energy management are integral components of the Internet of Energy (IoE) field. The integration of Software-Defined Networks (SDNs) and Blockchain (BC) technologies into the IoE domain offers potential benefits that have only been studied in the literature in a few works. In this paper, we investigate the state-of-art solutions that leverage both SDNs and blockchain within the realm of the IoE. We categorize these solutions based on the method of integrating SDN and BC into two categories. The first category is the blockchain for SDN, where blockchain enhances the SDN directly. The second category is blockchain and SDN, where both technologies are used to enhance the proposed solutions. We identify three distinct blockchain applications based on their usage: decentralizing the SDN control plane, serving as a decentralized platform, and improving security measures. Similarly, we observe that SDN serves as a performance enhancer, a substitute for traditional networking, and solely as a control and management framework. It is posited that integrating SDNs and blockchain into IoE leads to performance enhancements, improves security, enables decentralized operations, and eliminates single points of failure in the SDN control plane. Additionally, some unaddressed issues, such as energy efficiency, smart contract management, and scalability, are discussed as potential future directions., Comment: Presented in BCCA 2024: https://bcca-conference.org/2024/

This study investigates the potential of Large Language Models (LLMs), particularly GPT-4o, for Optical Character Recognition (OCR) in low-resource scripts such as Urdu, Albanian, and Tajik, with English serving as a benchmark. Using a meticulously curated dataset of 2,520 images incorporating controlled variations in text length, font size, background color, and blur, the research simulates diverse real-world challenges. Results emphasize the limitations of zero-shot LLM-based OCR, particularly for linguistically complex scripts, highlighting the need for annotated datasets and fine-tuned models. This work underscores the urgency of addressing accessibility gaps in text digitization, paving the way for inclusive and robust OCR solutions for underserved languages.

This summary of the second Terrestrial Very-Long-Baseline Atom Interferometry (TVLBAI) Workshop provides a comprehensive overview of our meeting held in London in April 2024, building on the initial discussions during the inaugural workshop held at CERN in March 2023. Like the summary of the first workshop, this document records a critical milestone for the international atom interferometry community. It documents our concerted efforts to evaluate progress, address emerging challenges, and refine strategic directions for future large-scale atom interferometry projects. Our commitment to collaboration is manifested by the integration of diverse expertise and the coordination of international resources, all aimed at advancing the frontiers of atom interferometry physics and technology, as set out in a Memorandum of Understanding signed by over 50 institutions., Comment: Summary of the second Terrestrial Very-Long-Baseline Atom Interferometry Workshop held at Imperial College London: https://indico.cern.ch/event/1369392/

Neuro-symbolic Artificial Intelligence (AI) models, blending neural networks with symbolic AI, have facilitated transparent reasoning and context understanding without the need for explicit rule-based programming. However, implementing such models in the Internet of Things (IoT) sensor nodes presents hurdles due to computational constraints and intricacies. In this work, for the first time, we propose a near-sensor neuro-symbolic AI computing accelerator named Neuro-Photonix for vision applications. Neuro-photonix processes neural dynamic computations on analog data while inherently supporting granularity-controllable convolution operations through the efficient use of photonic devices. Additionally, the creation of an innovative, low-cost ADC that works seamlessly with photonic technology removes the necessity for costly ADCs. Moreover, Neuro-Photonix facilitates the generation of HyperDimensional (HD) vectors for HD-based symbolic AI computing. This approach allows the proposed design to substantially diminish the energy consumption and latency of conversion, transmission, and processing within the established cloud-centric architecture and recently designed accelerators. Our device-to-architecture results show that Neuro-Photonix achieves 30 GOPS/W and reduces power consumption by a factor of 20.8 and 4.1 on average on neural dynamics compared to ASIC baselines and photonic accelerators while preserving accuracy., Comment: 12 pages, 15 figures

Non-verbal communication often comprises of semantically rich gestures that help convey the meaning of an utterance. Producing such semantic co-speech gestures has been a major challenge for the existing neural systems that can generate rhythmic beat gestures, but struggle to produce semantically meaningful gestures. Therefore, we present RAG-Gesture, a diffusion-based gesture generation approach that leverages Retrieval Augmented Generation (RAG) to produce natural-looking and semantically rich gestures. Our neuro-explicit gesture generation approach is designed to produce semantic gestures grounded in interpretable linguistic knowledge. We achieve this by using explicit domain knowledge to retrieve exemplar motions from a database of co-speech gestures. Once retrieved, we then inject these semantic exemplar gestures into our diffusion-based gesture generation pipeline using DDIM inversion and retrieval guidance at the inference time without any need of training. Further, we propose a control paradigm for guidance, that allows the users to modulate the amount of influence each retrieval insertion has over the generated sequence. Our comparative evaluations demonstrate the validity of our approach against recent gesture generation approaches. The reader is urged to explore the results on our project page., Comment: Preprint. Project page: https://vcai.mpi-inf.mpg.de/projects/RAG-Gesture/

Surface visualizations are essential in analyzing three-dimensional spatiotemporal phenomena. Given its ability to provide enhanced spatial perception and scene maneuverability, virtual reality (VR) is an essential medium for surface visualization and interaction tasks. Such tasks primarily rely on visual cues that require an unoccluded view of the surface region under consideration. Haptic force feedback is a tangible interaction modality that alleviates the reliance on visual-only cues by allowing a direct physical sensation of the surface. In this paper, we evaluate the use of a force-based haptic stylus compared to handheld VR controllers via a between-subjects user study involving fundamental interaction tasks performed on surface visualizations. Keeping a consistent visual design across both modalities, our study incorporates tasks that require the localization of the highest, lowest, and random points on surfaces; and tasks that focus on brushing curves on surfaces with varying complexity and occlusion levels. Our findings show that participants took longer to brush curves using the haptic modality but could draw smoother curves compared to the handheld controllers. In contrast, haptics was faster in point localization, but the accuracy depended on the visual cues and occlusions associated with the tasks. Finally, we discuss participant feedback on using haptic force feedback as a tangible input modality and share takeaways to help outline design strategies for using haptics-based tangible inputs for surface visualization and interaction tasks., Comment: arXiv admin note: substantial text overlap with arXiv:2408.04031

Given a budget on total model size, one must decide whether to train a single, large neural network or to combine the predictions of many smaller networks. We study this trade-off for ensembles of random-feature ridge regression models. We prove that when a fixed number of trainable parameters are partitioned among $K$ independently trained models, $K=1$ achieves optimal performance, provided the ridge parameter is optimally tuned. We then derive scaling laws which describe how the test risk of an ensemble of regression models decays with its total size. We identify conditions on the kernel and task eigenstructure under which ensembles can achieve near-optimal scaling laws. Training ensembles of deep convolutional neural networks on CIFAR-10 and a transformer architecture on C4, we find that a single large network outperforms any ensemble of networks with the same total number of parameters, provided the weight decay and feature-learning strength are tuned to their optimal values.

We present CALICO, a method to fine-tune Large Language Models (LLMs) to localize conversational agent training data from one language to another. For slots (named entities), CALICO supports three operations: verbatim copy, literal translation, and localization, i.e. generating slot values more appropriate in the target language, such as city and airport names located in countries where the language is spoken. Furthermore, we design an iterative filtering mechanism to discard noisy generated samples, which we show boosts the performance of the downstream conversational agent. To prove the effectiveness of CALICO, we build and release a new human-localized (HL) version of the MultiATIS++ travel information test set in 8 languages. Compared to the original human-translated (HT) version of the test set, we show that our new HL version is more challenging. We also show that CALICO out-performs state-of-the-art LINGUIST (which relies on literal slot translation out of context) both on the HT case, where CALICO generates more accurate slot translations, and on the HL case, where CALICO generates localized slots which are closer to the HL test set., Comment: Accepted to The 37th International Conference on Neural Information Processing Systems (NeurIPS 2023) December 10-16, 2023 - SyntheticData4ML Workshop, New Orleans, United States https://neurips.cc/virtual/2023/workshop/66540

The principles of data spaces for sovereign data exchange across trusted organizations have so far mainly been adopted in business-to-business settings, and recently scaled to cloud environments. Meanwhile, research organizations have established distributed research data infrastructures, respecting the principle that data must be FAIR, i.e., findable, accessible, interoperable and reusable. For mutual benefit of these two communities, the FAIR Data Spaces project aims to connect them towards the vision of a common, cloud-based data space for industry and research. Thus, the project establishes a common legal and ethical framework, common technical building blocks, and it demonstrates the orchestration of multiple building blocks in self-contained settings addressing a diverse range of use cases in domains including health, biodiversity, and engineering. This paper gives a summary of all demonstrators, ranging from research data infrastructures scaled to industry-ready cloud environments to work in progress on building bridges between operational business-to-business data spaces and research data infrastructures., Comment: Accepted at the 40th ACM/SIGAPP Symposium On Applied Computing

This paper introduces TemporalVLM, a video large language model capable of effective temporal reasoning and fine-grained understanding in long videos. At the core, our approach includes a visual encoder for mapping a long-term input video into features which are time-aware and contain both local and global cues. In particular, it first divides the input video into short-term clips, which are jointly encoded with their timestamps into time-sensitive local features. Next, the local features are passed through a bidirectional long short-term memory module for global feature aggregation. The extracted time-aware and multi-level features are important for accurate temporal reasoning and fine-grained understanding in long videos. Moreover, to facilitate the evaluation of TemporalVLM, we present a large-scale long video dataset of industry assembly processes, namely IndustryASM, which consists of videos recorded on factory floors with actions and timestamps annotated by industrial engineers for time and motion studies and temporal action segmentation evaluation. Finally, extensive experiments on datasets of long videos, including TimeIT and IndustryASM, show that TemporalVLM achieves superior performance than previous methods across temporal reasoning and fine-grained understanding tasks, namely dense video captioning, temporal video grounding, video highlight detection, and temporal action segmentation.

In this paper, the CHSH quantum game is extended to four players. This is achieved by exploring all possible 4-variable Boolean functions to identify those that yield a game scenario with a quantum advantage using a specific entangled state. Notably, two new four-player quantum games are presented. In one game, the optimal quantum strategy is achieved when players share a $GHZ$-state, breaking the traditional 10\% gain observed in 2 and 3 qubit CHSH games and achieving a 22.5\% gap. In the other game, players gain a greater advantage using a $W$-state as their quantum resource. Quantum games with other four-qubit entangled states are also explored. To demonstrate the results, these game scenarios are implemented on an online quantum computer, and the advantage of the respective quantum resource for each game is experimentally verified., Comment: 17 pages, 11 figures, 7 tables

Hyperdimensional computing (HDC) enables efficient data encoding and processing in high-dimensional space, benefiting machine learning and data analysis. However, underutilization of these spaces can lead to overfitting and reduced model reliability, especially in data-limited systems a critical issue in sectors like healthcare that demand robustness and consistent performance. We introduce BoostHD, an approach that applies boosting algorithms to partition the hyperdimensional space into subspaces, creating an ensemble of weak learners. By integrating boosting with HDC, BoostHD enhances performance and reliability beyond existing HDC methods. Our analysis highlights the importance of efficient utilization of hyperdimensional spaces for improved model performance. Experiments on healthcare datasets show that BoostHD outperforms state-of-the-art methods. On the WESAD dataset, it achieved an accuracy of 98.37%, surpassing Random Forest, XGBoost, and OnlineHD. BoostHD also demonstrated superior inference efficiency and stability, maintaining high accuracy under data imbalance and noise. In person-specific evaluations, it achieved an average accuracy of 96.19%, outperforming other models. By addressing the limitations of both boosting and HDC, BoostHD expands the applicability of HDC in critical domains where reliability and precision are paramount., Comment: Accepted to DATE 2025

Backscatter Communication (BackCom) technology has emerged as a promising paradigm for the Green Internet of Things (IoT) ecosystem, offering advantages such as low power consumption, cost-effectiveness, and ease of deployment. While traditional BackCom systems, such as RFID technology, have found widespread applications, the advent of ambient backscatter presents new opportunities for expanding applications and enhancing capabilities. Moreover, ongoing standardization efforts are actively focusing on BackCom technologies, positioning them as a potential solution to meet the near-zero power consumption and massive connectivity requirements of next-generation wireless systems. 0G networks have the potential to provide advanced solutions by leveraging BackCom technology to deliver ultra-low-power, ubiquitous connectivity for the expanding IoT ecosystem, supporting billions of devices with minimal energy consumption. This paper investigates the integration of BackCom and 0G networks to enhance the capabilities of traditional BackCom systems and enable Green IoT. We conduct an in-depth analysis of BackCom-enabled 0G networks, exploring their architecture and operational objectives, and also explore the Waste Factor (WF) metric for evaluating energy efficiency and minimizing energy waste within integrated systems. By examining both structural and operational aspects, we demonstrate how this synergy enhances the performance, scalability, and sustainability of next-generation wireless networks. Moreover, we highlight possible applications, open challenges, and future directions, offering valuable insights for guiding future research and practical implementations aimed at achieving large-scale, sustainable IoT deployments.

In response to the growing challenges of manual labor and efficiency in warehouse operations, Amazon has embarked on a significant transformation by incorporating robotics to assist with various tasks. While a substantial number of robots have been successfully deployed for tasks such as item transportation within warehouses, the complex process of object picking from shelves remains a significant challenge. This project addresses the issue by developing an innovative robotic system capable of autonomously fulfilling a simulated order by efficiently selecting specific items from shelves. A distinguishing feature of the proposed robotic system is its capacity to navigate the challenge of uncertain object positions within each bin of the shelf. The system is engineered to autonomously adapt its approach, employing strategies that enable it to efficiently locate and retrieve the desired items, even in the absence of pre-established knowledge about their placements., Comment: 5 PAGES

Enterprises are increasingly adopting serverless computing to enhance scalability, reduce costs, and improve efficiency. However, this shift introduces new responsibilities and necessitates a distinct set of skills for practitioners. This study aims to identify and organize the industry requirements for serverless practitioners by conducting a qualitative analysis of 141 job advertisements from seven countries. We developed comprehensive taxonomies of roles, responsibilities, and skills, categorizing 19 responsibilities into four themes: software development, infrastructure and operations, professional development and leadership, and software business. Additionally, we identified 28 hard skills mapped into seven themes and 32 soft skills mapped into eight themes, with the six most demanded soft skills being communication proficiency, continuous learning and adaptability, collaborative teamwork, problem-solving and analytical skills, leadership excellence, and project management. Our findings contribute to understanding the organizational structures and training requirements for effective serverless computing adoption.

Fetal brain tissue segmentation in magnetic resonance imaging (MRI) is a crucial tool that supports the understanding of neurodevelopment, yet it faces challenges due to the heterogeneity of data coming from different scanners and settings, and due to data scarcity. Recent approaches based on domain randomization, like SynthSeg, have shown a great potential for single source domain generalization, by simulating images with randomized contrast and image resolution from the label maps. In this work, we investigate how to maximize the out-of-domain (OOD) generalization potential of SynthSeg-based methods in fetal brain MRI. Specifically, when studying data generation, we demonstrate that the simple Gaussian mixture models used in SynthSeg enable more robust OOD generalization than physics-informed generation methods. We also investigate how intensity clustering can help create more faithful synthetic images, and observe that it is key to achieving a non-trivial OOD generalization capability when few label classes are available. Finally, by combining for the first time SynthSeg with modern fine-tuning approaches based on weight averaging, we show that fine-tuning a model pre-trained on synthetic data on a few real image-segmentation pairs in a new domain can lead to improvements in the target domain, but also in other domains. We summarize our findings as five key recommendations that we believe can guide practitioners who would like to develop SynthSeg-based approaches in other organs or modalities.

In this work, we significantly expand the web of T-dualities among heterotic NS5-brane theories with eight supercharges. This is achieved by introducing twists involving outer automorphisms of discrete gauge/flavor factors and tensor multiplet permutations along the compactification circle. We assemble field theory data that we propose as invariants across T-dual theories, comprised of twisted Coulomb branch dimensions, higher group structures and flavor symmetry ranks. Using this data, we establish a detailed field theory correspondence between singularities of the compactification space, the number five-branes in the theory, and the flavor symmetry factors. The twisted theories are realized via M-theory compactifications on non-compact genus-one fibered Calabi-Yau threefolds without section. This approach allows us to prove duality of twisted and (un-)twisted theories by leveraging M/F-theory duality and identifying inequivalent torus fibrations in the same geometry. We construct several new 5D theories, including a novel type of CHL-like twisted theory where the two M9 branes are identified. Using their field theory invariants, we also construct their dual theories., Comment: 62 pages, 2 figures, 12 tables

We establish a generalized quantum asymptotic equipartition property (AEP) beyond the i.i.d. framework where the random samples are drawn from two sets of quantum states. In particular, under suitable assumptions on the sets, we prove that all operationally relevant divergences converge to the quantum relative entropy between the sets. More specifically, both the smoothed min- and max-relative entropy approach the regularized relative entropy between the sets. Notably, the asymptotic limit has explicit convergence guarantees and can be efficiently estimated through convex optimization programs, despite the regularization, provided that the sets have efficient descriptions. We give four applications of this result: (i) The generalized AEP directly implies a new generalized quantum Stein's lemma for conducting quantum hypothesis testing between two sets of quantum states. (ii) We introduce a quantum version of adversarial hypothesis testing where the tester plays against an adversary who possesses internal quantum memory and controls the quantum device and show that the optimal error exponent is precisely characterized by a new notion of quantum channel divergence, named the minimum output channel divergence. (iii) We derive a relative entropy accumulation theorem stating that the smoothed min-relative entropy between two sequential processes of quantum channels can be lower bounded by the sum of the regularized minimum output channel divergences. (iv) We apply our generalized AEP to quantum resource theories and provide improved and efficient bounds for entanglement distillation, magic state distillation, and the entanglement cost of quantum states and channels. At a technical level, we establish new additivity and chain rule properties for the measured relative entropy which we expect will have more applications., Comment: comments are welcome

We introduce Agent K v1.0, an end-to-end autonomous data science agent designed to automate, optimise, and generalise across diverse data science tasks. Fully automated, Agent K v1.0 manages the entire data science life cycle by learning from experience. It leverages a highly flexible structured reasoning framework to enable it to dynamically process memory in a nested structure, effectively learning from accumulated experience stored to handle complex reasoning tasks. It optimises long- and short-term memory by selectively storing and retrieving key information, guiding future decisions based on environmental rewards. This iterative approach allows it to refine decisions without fine-tuning or backpropagation, achieving continuous improvement through experiential learning. We evaluate our agent's apabilities using Kaggle competitions as a case study. Following a fully automated protocol, Agent K v1.0 systematically addresses complex and multimodal data science tasks, employing Bayesian optimisation for hyperparameter tuning and feature engineering. Our new evaluation framework rigorously assesses Agent K v1.0's end-to-end capabilities to generate and send submissions starting from a Kaggle competition URL. Results demonstrate that Agent K v1.0 achieves a 92.5\% success rate across tasks, spanning tabular, computer vision, NLP, and multimodal domains. When benchmarking against 5,856 human Kaggle competitors by calculating Elo-MMR scores for each, Agent K v1.0 ranks in the top 38\%, demonstrating an overall skill level comparable to Expert-level users. Notably, its Elo-MMR score falls between the first and third quartiles of scores achieved by human Grandmasters. Furthermore, our results indicate that Agent K v1.0 has reached a performance level equivalent to Kaggle Grandmaster, with a record of 6 gold, 3 silver, and 7 bronze medals, as defined by Kaggle's progression system.

We consider the problem of learning the Hamiltonian of a quantum system from estimates of Gibbs-state expectation values. Various methods for achieving this task were proposed recently, both from a practical and theoretical point of view. On the one hand, some practical algorithms have been implemented and used to analyze experimental data but these algorithms often lack correctness guarantees or fail to scale to large systems. On the other hand, theoretical algorithms with provable asymptotic efficiency guarantees have been proposed, but they seem challenging to implement. Recently, a semidefinite family of Hamiltonian learning algorithms was proposed which was numerically demonstrated to scale well into the 100-qubit regime, but provided no provable accuracy guarantees. We build on this work in two ways, by extending it to provide certified a posteriori lower and upper bounds on the parameters to be learned, and by proving a priori convergence in the special case where the Hamiltonian is commuting.

Objective: This review aims to analyze the application of natural language processing (NLP) techniques in cancer research using electronic health records (EHRs) and clinical notes. This review addresses gaps in the existing literature by providing a broader perspective than previous studies focused on specific cancer types or applications. Methods: A comprehensive literature search was conducted using the Scopus database, identifying 94 relevant studies published between 2019 and 2024. Data extraction included study characteristics, cancer types, NLP methodologies, dataset information, performance metrics, challenges, and future directions. Studies were categorized based on cancer types and NLP applications. Results: The results showed a growing trend in NLP applications for cancer research, with breast, lung, and colorectal cancers being the most studied. Information extraction and text classification emerged as predominant NLP tasks. A shift from rule-based to advanced machine learning techniques, particularly transformer-based models, was observed. The Dataset sizes used in existing studies varied widely. Key challenges included the limited generalizability of proposed solutions and the need for improved integration into clinical workflows. Conclusion: NLP techniques show significant potential in analyzing EHRs and clinical notes for cancer research. However, future work should focus on improving model generalizability, enhancing robustness in handling complex clinical language, and expanding applications to understudied cancer types. Integration of NLP tools into clinical practice and addressing ethical considerations remain crucial for utilizing the full potential of NLP in enhancing cancer diagnosis, treatment, and patient outcomes.

This paper introduces Ali-AUG, a novel single-step diffusion model for efficient labeled data augmentation in industrial applications. Our method addresses the challenge of limited labeled data by generating synthetic, labeled images with precise feature insertion. Ali-AUG utilizes a stable diffusion architecture enhanced with skip connections and LoRA modules to efficiently integrate masks and images, ensuring accurate feature placement without affecting unrelated image content. Experimental validation across various industrial datasets demonstrates Ali-AUG's superiority in generating high-quality, defect-enhanced images while maintaining rapid single-step inference. By offering precise control over feature insertion and minimizing required training steps, our technique significantly enhances data augmentation capabilities, providing a powerful tool for improving the performance of deep learning models in scenarios with limited labeled data. Ali-AUG is especially useful for use cases like defective product image generation to train AI-based models to improve their ability to detect defects in manufacturing processes. Using different data preparation strategies, including Classification Accuracy Score (CAS) and Naive Augmentation Score (NAS), we show that Ali-AUG improves model performance by 31% compared to other augmentation methods and by 45% compared to models without data augmentation. Notably, Ali-AUG reduces training time by 32% and supports both paired and unpaired datasets, enhancing flexibility in data preparation.

Advancements in light modulator technology have been driving discoveries and progress across various fields. The problem of large-scale coherent optical control of atomic quantum systems-including cold atoms, ions, and solid-state color centers-presents among the most stringent requirements. This motivates a new generation of high-speed large-scale modulator technology with the following requirements: (R1) operation at a design wavelength of choice in the visible (VIS) to near-infrared (NIR) spectrum, (R2) a scalable technology with a high channel density (> 100mm-2 ), (R3) a high modulation speed (> 100MHz), and (R4) a high extinction ratio (> 20 dB). To fulfill these requirements, we introduce a modulator technology based on piezoelectrically actuated silicon nitride resonant waveguide gratings fabricated on 200mm diameter silicon wafers with CMOS compatible processes. We present a proof-of-concept device with 4 x 4 individually addressable 50 {mu}m x 50 {mu}m pixels or channels, each containing a resonant waveguide grating with a ~ 780 nm design wavelength, supporting > 100MHz modulation speeds, and a spectral response with > 20 dB extinction., Comment: 9 pages + 2 pages supplementary

We introduce QICS (Quantum Information Conic Solver), an open-source primal-dual interior point solver fully implemented in Python, which is focused on solving optimization problems arising in quantum information theory. QICS has the ability to solve optimization problems involving the quantum relative entropy, noncommutative perspectives of operator convex functions, and related functions. It also includes an efficient semidefinite programming solver which exploits sparsity, as well as support for Hermitian matrices. QICS is also currently supported by the Python optimization modelling software PICOS. This paper aims to document the implementation details of the algorithm and cone oracles used in QICS, and serve as a reference guide for the software. Additionally, we showcase extensive numerical experiments which demonstrate that QICS outperforms state-of-the-art quantum relative entropy programming solvers, and has comparable performance to state-of-the-art semidefinite programming solvers., Comment: 35 pages, 4 figures, 5 tables. v2: Minor changes to Section 5

Medical report generation is the task of automatically writing radiology reports for chest X-ray images. Manually composing these reports is a time-consuming process that is also prone to human errors. Generating medical reports can therefore help reduce the burden on radiologists. In other words, we can promote greater clinical automation in the medical domain. In this work, we propose a new framework leveraging vision-enabled Large Language Models (LLM) for the task of medical report generation. We introduce a lightweight solution that achieves better or comparative performance as compared to previous solutions on the task of medical report generation. We conduct extensive experiments exploring different model sizes and enhancement approaches, such as prefix tuning to improve the text generation abilities of the LLMs. We evaluate our approach on a prominent large-scale radiology report dataset - MIMIC-CXR. Our results demonstrate the capability of our resource-efficient framework to generate patient-specific reports with strong medical contextual understanding and high precision.

Neuro-symbolic artificial intelligence (AI) excels at learning from noisy and generalized patterns, conducting logical inferences, and providing interpretable reasoning. Comprising a 'neuro' component for feature extraction and a 'symbolic' component for decision-making, neuro-symbolic AI has yet to fully benefit from efficient hardware accelerators. Additionally, current hardware struggles to accommodate applications requiring dynamic resource allocation between these two components. To address these challenges-and mitigate the typical data-transfer bottleneck of classical Von Neumann architectures-we propose a ferroelectric charge-domain compute-in-memory (CiM) array as the foundational processing element for neuro-symbolic AI. This array seamlessly handles both the critical multiply-accumulate (MAC) operations of the 'neuro' workload and the parallel associative search operations of the 'symbolic' workload. To enable this approach, we introduce an innovative 1FeFET-1C cell, combining a ferroelectric field-effect transistor (FeFET) with a capacitor. This design, overcomes the destructive sensing limitations of DRAM in CiM applications, while capable of capitalizing decades of DRAM expertise with a similar cell structure as DRAM, achieves high immunity against FeFET variation-crucial for neuro-symbolic AI-and demonstrates superior energy efficiency. The functionalities of our design have been successfully validated through SPICE simulations and prototype fabrication and testing. Our hardware platform has been benchmarked in executing typical neuro-symbolic AI reasoning tasks, showing over 2x improvement in latency and 1000x improvement in energy efficiency compared to GPU-based implementations.

We present, for the first time, power splitters with multiple channel configurations in one-dimensional grating waveguides (1DGWs) that maintain crystal lattice-sensitive Bloch mode profiles without perturbation across all output channels, all within an ultra-miniaturized footprint of just 2.1 x 2.2 {\mu}m^2. This novel capability reduces the need for transition regions, simplifies multi-channel configurations of 1DGWs, and maximizes the effective use of chip area. The pixelated metamaterial approach, integrated with a time-domain heuristic algorithm, is utilized to concurrently achieve broadband operation, optimized dispersion control, and minimal loss. We experimentally demonstrate that our 1x2 and 1x3 metasplitters achieve average minimum losses per channel of 3.80 dB and 5.36 dB, respectively, just 0.80 dB and 0.59 dB above ideal splitting. The measurements for both designs demonstrate a 1 dB bandwidth of 15 nm, with excellent uniformity across all output channels. These versatile metasplitter designs can serve as fundamental building blocks for ultrahigh-bandwidth, densely integrated photonic circuits and in scenarios where slow light is essential., Comment: 9 pages, 3 figures, 7 pages supplementary information