2787 papers
This study presents a lightweight Faster R-CNN framework enhanced with a standalone BiFPN that achieves high-accuracy detection and counting of tuberculosis bacilli in Ziehl-Neelsen smear microscopy, offering a computationally efficient AI solution to accelerate diagnosis and reduce variability in tuberculosis control.
This paper introduces an ultrafast photonic processor that leverages the temporal speckle of broadband light to natively encode probabilistic weights, thereby overcoming the computational bottlenecks of digital Bayesian neural networks and enabling high-speed, uncertainty-aware AI capable of simultaneous classification and out-of-distribution detection.
This paper proposes a novel framework combining Hamiltonian Full Node Coverage Graph Attention Networks with Fuzzy C-Means superpixel graph learning to achieve highly accurate and reliable classification of banana leaf diseases by effectively modeling both local and global spatial dependencies.
The paper introduces Failure-Aware Adaptive Inference (FAAI), a framework that leverages reusable failure knowledge through adaptive memory retrieval, risk estimation, and proactive self-correction to significantly improve the reliability and success rates of long-horizon language model agents without requiring additional model training.
This systematic review evaluates the role of semantic interoperability in autonomous data ecosystems, highlighting the adoption of ontologies and knowledge graphs while identifying critical challenges such as semantic heterogeneity, scalability, and the need for standardized evaluation in industrial applications.
This paper introduces ExecuGraph, a configurable framework demonstrating that execution feedback is the primary driver of improved code generation accuracy in large language models, while showing that adding multi-agent role decomposition provides no measurable benefit over single-agent retry loops despite significantly higher computational costs.
This paper presents a comprehensive review of 32 studies to propose a five-layer Explainable AI framework that integrates technical methods, human factors, and governance instruments to address the transparency and trust challenges of deploying "black-box" AI in critical IT systems.
This paper presents a transparent proof-of-concept for a compact teacher-student prototype that explicitly models the tension between factual reliability and affective sensitivity in dialogue, demonstrating successful in-sample compression while highlighting the system's failure to improve truthfulness or empathy and proposing a concrete redesign path for future multi-objective studies.
This review of 247 studies on Intelligent Product Conceptual Design reveals that while the field is shifting toward hybrid AI architectures, persistent trade-offs between interpretability, scalability, and human agency necessitate a future research agenda focused on distributed design intelligence, standardized benchmarks, and deeper industrial integration.
This paper presents a systematic mapping study of 46 language models developed for Portuguese, offering a comprehensive overview of their technical characteristics, evolutionary relationships, and current research gaps to guide future advancements in the field.