Abstract

Temporal Graph Neural Networks (TGNNs) are increasingly used in high-stakes domains, such as financial forecasting, recommendation systems, and fraud detection. However, their susceptibility to poisoning attacks poses a critical security risk. We introduce LORETTA (Low Resource Twophase Temporal Attack), a novel adversarial framework on Continuous-Time Dynamic Graphs, which degrades TGNN performance by an average of 29.47% across 4 widely benchmark datasets and 4 State-of-the-Art (SotA) models. LORETTA operates through a two-stage approach: (1) sparsify the graph by removing high-impact edges using any of the 16 tested temporal importance metrics, (2) strategically replace removed edges with adversarial negatives via LORETTA’s novel degree-preserving negative sampling algorithm. Our plug-and-play design eliminates the need for expensive surrogate models while adhering to realistic unnoticeability constraints. LORETTA degrades performance by upto 42.0% on MOOC, 31.5% on Wikipedia, 28.8% on UCI, and 15.6%on Enron. LORETTA outperforms 11 attack baselines, remains undetectable to 4 leading anomaly detection systems, and is robust to 4 SotA adversarial defense training methods, establishing its effectiveness, unnoticeability, and robustness.

Abstract

Traditional solar forecasting models are based on several years of site-specific historical irradiance data, often spanning five or more years, which are unavailable for newer photovoltaic farms. As renewable energy is highly intermittent, building accurate solar irradiance forecasting systems that are data-efficient is essential for efficient grid management and enabling the ongoing proliferation of solar energy, which is crucial to achieve the United Nations' net zero goals. In this work, we propose SPIRIT, a novel framework leveraging foundation models for solar irradiance forecasting, making it applicable to newer solar installations. Our approach outperforms state-of-the-art models in zero-shot transfer learning by upto 70%, enabling effective performance at new locations without relying on any past data. Further improvements in performance are achieved through fine-tuning, as more location-specific data becomes available. These findings are supported by statistical significance, further validating our approach. By dramatically reducing the forecasting setup timeline, SPIRIT accelerates solar farm deployment in all potential global sites, most of which lack historical data, thereby democratizing access to clean energy and enabling participation in the renewable energy transition.

Abstract

In recent times, denoising diffusion probabilistic models (DPMs) have proven to show significant success in medical image generation and denoising, while also serving as powerful representation learners for downstream tasks such as segmentation. However, their effectiveness in segmentation is limited by the need for detailed pixel-wise annotations, which are expensive, time-consuming, and require expert knowledge—a significant bottleneck in real-world clinical applications. In order to mitigate this limitation of label-efficiency, we propose a fast and efficient model named FastTextDiff, a diffusion-based segmentation model that integrates medical text annotations to enhance semantic representations. Our approach leverages ModernBERT, a transformer-based language model capable of processing long medical text sequences, to establish a strong connection between textual annotations and semantic meaning in medical imaging. ModernBERT can efficiently encode clinical knowledge for directing segmentation tasks since it has been trained on both MIMIC-III and MIMIC-IV. Label-efficient segmentation with enhanced performance is made possible by cross-modal attention processes, which enable smooth interaction between visual and textual modalities. This study validates ModernBERT as a quick and scalable substitute for Clinical BioBERT in diffusion-based segmentation pipelines and demonstrates the promise of multi-modal techniques for medical image analysis. By replacing Clinical BioBERT with ModernBERT, our model benefits from Flash Attention 2 for memory-efficient training, an alternating attention mechanism for computational efficiency, and a training corpus of 2 trillion tokens, significantly improving text-based medical image segmentation. Our experiments demonstrate that FastTextDiff achieves better segmentation performance and faster training compared to traditional diffusion-based models. We release this trained model on Hugging Face and GitHub.

Abstract

Natural fibre composites (NFCs) have gained prominence as sustainable, cost-effective alternatives to synthetic fibre reinforcements in masonry structural retrofitting. This paper reviews the extraction, treatment, and fabrication of natural fibres such as jute, flax, sisal, hemp, and basalt into Fibre Reinforced Polymer (FRP) and Textile Reinforced Mortar (TRM) systems suited for masonry appli-cations. Experimental and field studies demonstrate that natural FRP composites improve tensile strength, ductility, and seismic resilience of unreinforced masonry walls, while natural TRM composites offer enhanced environmental compatibil-ity, fire resistance, and vapour permeability, which are essential for heritage ma-sonry preservation. The review assesses mechanical performance, durability chal-lenges related to moisture and ultraviolet (UV) exposure, and sustainability bene-fits, including reduced embodied energy and biodegradability. Practical case stud-ies confirm the feasibility of locally sourced natural fibre retrofits in seismic zones. Challenges such as variability in fibre properties and the need for stand-ardized design codes are discussed alongside future research directions focusing on hybrid composites, bio-based matrices, and sensor integration for structural health monitoring. This review consolidates current advances and prospects, positioning NFCs in form of FRP and TRM composites as viable, eco-friendly retrofit technologies for resilient and sustainable masonry infrastructure.

Abstract

Bharatanatyam is one of the most celebrated classical dance forms in India. It is performed and enjoyed by art enthusiasts all over the world. It comprises expressions, hand gestures, footwork, rhythmic patterns, eye, neck and head movements, lyrics, and music. These basic components are integrated and woven into a beautifully crafted visual story. However, beneath this creative visual craft lies a structured compilation of technical, mathematical, computational, and modular aspects. In this paper, we provide a database schema to store the artistic components in a structured relational database model. By providing a database schema, we provide a layered approach to store Bharatanatyam choreographies. This layered approach helps in building choreographies by working individually on different modules of the art form like the foot work, hand movement, etc, and consolidating them to form and store sequences. Bharatanatyam choreographies are a collection of sequences. In this paper, we discuss how complete choreographies are modelled through the logical sequencing of these components. Our paper lays the foundation for bridging traditional knowledge of the performing arts with technologies to help in digital archiving, cultural analytics, and choreography suggestions.

Abstract

Non-Structural Elements (NSEs) such as building contents, appendages to buildings, utilities and services are highly vulnerable during earthquakes and can pose significant risks if not properly secured to the supporting structure. Consequences of unsecured NSEs in the aftermath of earthquakes are significant losses to human life, property, and functionality of buildings. This is because, NSEs are susceptible to sliding, rocking, toppling, overturning, stretching, shortening, and shearing. These movements are compounded by: (a) NSE-structure anchorage conditions, (b) geometry, weight, and friction with the supporting surface and NSE, and (c) NSE types. Recent earthquakes in Turkey and Myanmar also witnessed damage in NSEs, reiterating the criticality. In general, earthquake resistance of NSEs can be improved by 3 broad mitigation approaches recommended in literature. They are: (i) non-engineered (based on judgement) (ii) pre-engineered (based on calculations, experiments, and experience from past earthquakes), and (iii) engineered (based on detailed design recommended in design standards). These approaches typically employ anchor bolts, braces, slack chains, snubbers, flexible connections, and shock absorbers tailored to the specific type and performance requirements of NSE. This paper presents a detailed discussion of commonly adopted methods that help improve earthquake resistance of NSEs and enhance overall safety of buildings in urban areas.

Abstract

Accurate segmentation of medical images is challenging due to unclear lesion boundaries and mask variability. We introduce Segmentation Schödinger Bridge (SSB), the first application of Schödinger Bridge for ambiguous medical image segmentation, modelling joint image-mask dynamics to enhance performance. SSB preserves structural integrity, delineates unclear boundaries without additional guidance, and maintains diversity using a novel loss function. We further propose the Diversity Divergence Index () to quantify inter-rater variability, capturing both diversity and consensus. SSB achieves state-of-the-art performance on LIDC-IDRI, COCA, and RACER (in-house) datasets.

Abstract

Extracting work from a physical system is one of the cornerstones of quantum thermodynamics. The extractable work, as quantified by ergotropy, necessitates a complete description of the quantum system. This is significantly more challenging when the state of the underlying system is unknown, as quantum tomography is extremely inefficient. In this article, we analyze the number of samples of the unknown state required to extract work. With only a single copy of an unknown state, we prove that ergotropy approaches zero in the asymptotic limit, rendering work extraction nearly impossible. In contrast, when multiple copies are available, we quantify the sample complexity required to estimate extractable work, establishing a scaling relationship that balances the desired accuracy with success probability. Our work develops a sample-efficient protocol to assess the utility of unknown states as quantum batteries and opens avenues for estimating thermodynamic quantities using near-term quantum computers.

Abstract

Curating high-quality, domain-specific datasets is a major bottleneck for deploying robust vision systems, requiring complex trade-offs between data quality, diversity, and cost when researching vast, unlabeled data lakes. We introduce Labeling Copilot, the first data curation deep research agent for computer vision. A central orchestrator agent, powered by a large multimodal language model, uses multi-step reasoning to execute specialized tools across three core capabilities: (1) Calibrated Discovery sources relevant, in-distribution data from large repositories; (2) Controllable Synthesis generates novel data for rare scenarios with robust filtering; and (3) Consensus Annotation produces accurate labels by orchestrating multiple foundation models via a novel consensus mechanism incorporating non-maximum suppression and voting. Our large-scale validation proves the effectiveness of Labeling Copilot's components. The Consensus Annotation module excels at object discovery: on the dense COCO dataset, it averages 14.2 candidate proposals per image-nearly double the 7.4 ground-truth objects-achieving a final annotation mAP of 37.1%. On the web-scale Open Images dataset, it navigated extreme class imbalance to discover 903 new bounding box categories, expanding its capability to over 1500 total. Concurrently, our Calibrated Discovery tool, tested at a 10-million sample scale, features an active learning strategy that is up to 40x more computationally efficient than alternatives with equivalent sample efficiency. These experiments validate that an agentic workflow with optimized, scalable tools provides a robust foundation for curating industrial-scale datasets.

Abstract

Excessive groundwater extraction has led to a significant decline in water tables, highlighting the need for continuous monitoring. This study presents an Internet of Things (IoT)-based approach for real-time estimation of groundwater levels in borewells. In addition to water level measurements, parameters such as motor current and rainfall are tracked to evaluate their influence on groundwater dynamics. To assess practical feasibility, five monitoring nodes were deployed across a small educational campus in Hyderabad, India. Data collected over a two-month period indicate that the proposed low-cost system is both reliable and effective in real-world conditions, while also revealing interesting correlations between groundwater level variations and the monitored parameters.