Wei-Xuan Hong

Undergraduate Research Assistant @Integrate Uw-Madison

Madison, WI, US

I contribute to research on human movement by developing GPU-based pipelines that integrate computer vision and machine learning models (YOLO-Pose, MiDaS/Video-Depth-Anything, Mask2Former) for detection, pose estimation, depth, and segmentation in video data. My work involves analyzing human trajectories, proximity, and intent in complex environments, preparing datasets, evaluating model performance.