3031 papers
This paper proposes a novel educational strategy that integrates Large Language Models and deep learning to create an Intelligent Virtual Friend Doll, enabling objective, continuous, and personalized assessment of educational progress for children with Autism Spectrum Disorder through the analysis of multimodal data.
This paper introduces vProjTutor, an LLM-agent-based system that integrates with tools like Telegram, Trello, and GitHub to support student teams through natural language interactions for project planning and monitoring, demonstrating high usability and task completion rates in a user study.
The paper introduces Asha, a neurosymbolic clinical agent architecture that significantly enhances medical accuracy and conversational safety in language models by integrating symbolic scoring, retrieval grounding, and falsification, achieving superior performance on professional benchmarks compared to frontier models and commercial tools.
This study reveals that lightweight intrusion detection models for IIoT networks suffer from severe cross-domain generalization failures due to reliance on non-transferable port-category shortcuts and evaluation biases, underscoring the critical need for cross-network testing under realistic class distributions to ensure deployment readiness.
This paper proposes an energy-efficient, context-aware multimodal inference framework for resource-constrained edge devices that dynamically activates encoders and adjusts computation paths based on real-time conditions, achieving significant reductions in energy consumption and latency while maintaining high accuracy on platforms like NVIDIA Jetson Nano and Raspberry Pi 5.
This paper introduces ViHistScriptBench, a lightweight benchmark and evaluation pipeline designed to advance the classification of Vietnamese historical scripts and document types by providing a curated corpus, defined tasks, and baseline models to address challenges posed by limited data and complex calligraphy.
This paper proposes a rule-engine and evidence-indexing methodology that bridges the semantic gap between audit requirements and software architecture by creating a classification model, mapping matrix, and verification algorithm to embed executable controls early in the design phase, thereby reducing remediation costs and enabling audit-ready digital infrastructure.
TrustChain leverages blockchain technology to revolutionize government tender management by ensuring transparency, immutability, and accountability through decentralized record-keeping and milestone-based payments, thereby reducing corruption and fostering fair competition.
This paper addresses the fragmentation in Augmented Workplace design by conducting a PRISMA-driven systematic review of 122 studies to consolidate 151 guidelines into a structured, transferable knowledge base that supports the systematic development and deployment of AR systems in professional environments.
This paper proposes a hybrid NSGA-II algorithm (HENSGA-DRL) that integrates a critical-path graph-guided deep reinforcement learning operator and a multi-metric state-aware hierarchical intervention strategy to overcome the structural perception and stagnation limitations of existing evolutionary methods in solving the multi-objective distributed flexible job shop scheduling problem.