2661 papers
This study demonstrates that fairness-aware machine learning models, specifically those employing bias mitigation techniques like Disparate Impact Remover and adversarial debiasing, can significantly reduce sex-related disparities in Alzheimer's disease classification using ADNI data while maintaining high predictive accuracy and enhancing the interpretability of clinical and gender-related determinants.
This paper introduces an automated, contamination-controlled benchmarking pipeline that generates multiple-choice questions from private 3D oncology imaging and radiology reports to rigorously evaluate Vision-Language Models, revealing that current models often rely on text cues or dataset familiarity rather than genuine visual perception.
This paper introduces DualAttentionNet, a lightweight deep learning framework that enhances traditional EEGNet by incorporating dual channel-attention blocks to achieve a state-of-the-art 92.36% accuracy in multi-class EEG digit decoding while ensuring robustness, interpretability, and computational efficiency.
This paper presents a systematic literature review (2020–2025) of AI-based Android malware detection systems, analyzing their vulnerabilities to adversarial attacks, evaluating existing defense strategies, and identifying critical gaps and future research directions for achieving robust and trustworthy security.
LC-MSCKF improves Visual-Inertial Odometry accuracy by employing a hybrid two-stage uncertainty calibration strategy that combines a lightweight learned branch for anticipating degradation with an innovation-feedback branch for bounded corrective inflation, significantly outperforming fixed and single-branch adaptive baselines across diverse benchmarks.
This study demonstrates that automated language analysis of clinical interviews using transformer-based models and ensemble strategies can effectively detect major depressive disorder in Cantonese speakers, significantly improving processing efficiency while highlighting the importance of transcription quality and conversational context for optimal performance.
This paper presents GMD-YOLO26, a lightweight UAV-based detector that employs a synergistic three-stage pipeline comprising Gated Convolutional Feature Enhancement, Multi-Scale Dilated Attention, and Dynamic Cross-scale Feature Aggregation to effectively address small-object detection challenges while achieving a favorable accuracy-efficiency trade-off on the VisDrone2019 benchmark.
This paper presents a locally deployable, knowledge distillation-based 4B language model system that effectively automates the standardized quality control of Chinese radiology reports, achieving high precision and significantly improved inference speed compared to larger teacher models.
This paper proposes a construct validity framework that conceptualizes ethical AI as a context-dependent, higher-order construct comprising fairness, transparency, and explainability, offering a theoretically grounded pathway to develop auditable measurement standards for regulatory compliance and empirical research.
This paper presents and empirically evaluates a Protocol-Governed Computing architecture that separates activity autonomy from authorization authority, demonstrating how human oversight can govern AI-driven software engineering through explicit admission, promotion, and sealing protocols while identifying specific areas where guard mechanisms require refinement.