2786 papers
This survey reviews semantic planning for collaborative robots by formalizing the challenge as a constrained grounding problem, proposing a taxonomy of architectures that balance capability with inspectability, and outlining a roadmap for enhancing reliability through improved uncertainty estimation, plan verification, and recovery mechanisms.
MECA-Net is a lightweight, real-time YOLO11n-based detector that enhances small traffic object detection in UAV imagery by integrating multi-scale edge-aware and coordinate-aware attention mechanisms, achieving significant mAP improvements on the VisDrone2019-DET dataset while maintaining high inference speeds.
This paper proposes a sampling-based data pipeline that leverages the stable Zipf distribution of game logs to efficiently identify and route dominant "hot keys" without full-stream monitoring, thereby achieving a 209.7% throughput improvement and significantly reduced CPU utilization compared to existing load-balancing solutions.
The paper proposes SKD-Net, a lightweight knowledge distillation framework featuring MGA, FSD, and ASD modules to achieve accurate and efficient coronary artery segmentation in X-ray angiography by transferring structural and semantic knowledge from a powerful teacher network while overcoming limitations in long-range relationship capture and boundary preservation.
This paper proposes a cross-device milling force prediction method that combines self-supervised pretraining, domain adaptation, and few-label fine-tuning to effectively transfer vibration-force mapping across devices despite dynamic response differences and limited target-domain labels, achieving high accuracy with minimal labeled data.
This paper exposes the illusion of high accuracy in medical speech-text symptom detection caused by sentence memorization in standard dataset splits, and establishes a rigorous phrase-disjoint evaluation baseline using the Adaptive Gated Cross-Modal Fusion (AGCF) model to reveal honest performance metrics, per-class fairness risks, and the limited utility of sentiment features while demonstrating significant efficiency gains through LoRA.
This paper proposes and evaluates an explainable IoT-driven framework that integrates interpretable AI models to enhance transparency, accountability, and real-time decision-making reliability in safety-critical autonomous vehicles.
This exploratory study of 37 upper secondary students reveals that while participants could not significantly distinguish between human and ChatGPT-generated programming solutions, their acceptance of a 3D avatar tutor was unaffected by prior programming experience or avatar familiarity, highlighting both optimism for and skepticism about integrating generative AI into education.
This paper proposes a hierarchical reinforcement learning-based framework that decomposes mural restoration into coordinated high-level region selection and low-level adaptive operations, utilizing reversible feature constraints to effectively balance noise suppression, structural fidelity, and conservative color recovery in complex degraded murals.
This paper introduces QTR-Net, a unified quantum-assisted and trust-aware Transformer framework for autonomous navigation that integrates trajectory prediction, dynamic trust mapping, and hybrid quantum-classical path optimization to achieve robust, adaptive control in complex traffic environments.