Session 4

[OPTM4-04] Human Pose Estimation from Noisy Point Cloud Using Unsupervised Learning

*Chaitali Bhattacharyya¹, Sungho Kim¹ (1. Yeungnam University)

3D computer vision, vital in industries like animation and healthcare, faces challenges with noisy point cloud data from depth cameras, affecting accuracy in human pose estimation. Our paper presents a novel method to accurately estimate poses from such data, enhancing applications in various fields. This breakthrough addresses key issues in 3D pose analysis, offering a reliable solution for diverse industries.