High-Dimensional Online Change Point Detection with Adaptive Thresholding and Interpretability
本文提出一种基于Sliced Wasserstein距离的在线变化点检测框架，通过自适应阈值调整和提高可解释性来解决高维数据中的计算挑战和实用性限制。
0 citationsRead paper
Selected work
Representative Papers
Train, Test, Re-evaluate: Schedule-Sensitive Evaluation of Generative Data for Hand Detection
This study addresses the degradation in hand detection performance caused by protective gear such as gloves and jewelry in occupational safety scenarios. To mitigate this issue, the authors propose a generative image inpainting–based synthetic data augmentation method that automatically overlays accessories onto hand regions of real images to simulate distribution shifts. Integrating this approach with the YOLOv8n detector, they devise a schedule-sensitive multi-stage fine-tuning strategy and rigorously evaluate its efficacy using statistical significance tests. Experimental results demonstrate that the three-stage training protocol achieves competitive mAP@0.5:0.95 on standard benchmarks, substantially improving detection robustness in gloved-hand scenarios, effectively narrowing the out-of-distribution generalization gap, and revealing that the utility of synthetic data is highly dependent on the training schedule.
0 citationsRead paper
Semantic Robustness Probing via Inpainting: An Interactive Tool for Safety-Critical Object Detection
Existing object detection models lack robustness evaluation against semantic-level perturbations in safety-critical scenarios, as conventional pixel-level perturbations fail to capture real-world semantic variations. To address this gap, this work proposes SemProbe—the first framework integrating controllable diffusion inpainting with safety assessment—enabling users to interactively design domain-relevant semantic perturbations via masking, automatically generate inpainted samples, and quantify model performance degradation through inference. The system supports batch experimentation, parallel workflow management, and structured logging to ensure traceable semantic robustness validation. Evaluated on a circular saw hand-detection task under insurance-guided testing criteria, SemProbe successfully generates effective semantic probes that expose critical model vulnerabilities under safety-relevant conditions.
0 citationsRead paper
Recent publications
Latest Papers
High-Dimensional Online Change Point Detection with Adaptive Thresholding and Interpretability
本文提出一种基于Sliced Wasserstein距离的在线变化点检测框架，通过自适应阈值调整和提高可解释性来解决高维数据中的计算挑战和实用性限制。
0 citationsRead paper
Train, Test, Re-evaluate: Schedule-Sensitive Evaluation of Generative Data for Hand Detection
This study addresses the degradation in hand detection performance caused by protective gear such as gloves and jewelry in occupational safety scenarios. To mitigate this issue, the authors propose a generative image inpainting–based synthetic data augmentation method that automatically overlays accessories onto hand regions of real images to simulate distribution shifts. Integrating this approach with the YOLOv8n detector, they devise a schedule-sensitive multi-stage fine-tuning strategy and rigorously evaluate its efficacy using statistical significance tests. Experimental results demonstrate that the three-stage training protocol achieves competitive mAP@0.5:0.95 on standard benchmarks, substantially improving detection robustness in gloved-hand scenarios, effectively narrowing the out-of-distribution generalization gap, and revealing that the utility of synthetic data is highly dependent on the training schedule.
0 citationsRead paper
Semantic Robustness Probing via Inpainting: An Interactive Tool for Safety-Critical Object Detection
Existing object detection models lack robustness evaluation against semantic-level perturbations in safety-critical scenarios, as conventional pixel-level perturbations fail to capture real-world semantic variations. To address this gap, this work proposes SemProbe—the first framework integrating controllable diffusion inpainting with safety assessment—enabling users to interactively design domain-relevant semantic perturbations via masking, automatically generate inpainted samples, and quantify model performance degradation through inference. The system supports batch experimentation, parallel workflow management, and structured logging to ensure traceable semantic robustness validation. Evaluated on a circular saw hand-detection task under insurance-guided testing criteria, SemProbe successfully generates effective semantic probes that expose critical model vulnerabilities under safety-relevant conditions.
0 citationsRead paper