C
Chengde Wan
Researcher at ETH Zurich
Publications - 17
Citations - 1178
Chengde Wan is an academic researcher from ETH Zurich. The author has contributed to research in topics: Pose & Depth map. The author has an hindex of 11, co-authored 16 publications receiving 897 citations. Previous affiliations of Chengde Wan include Facebook & Beijing University of Posts and Telecommunications.
Papers
More filters
Proceedings ArticleDOI
Deep Learning on Lie Groups for Skeleton-Based Action Recognition
TL;DR: The Lie group structure is incorporated into a deep network architecture to learn more appropriate Lie group features for 3D action recognition and a logarithm mapping layer is proposed to map the resulting manifold data into a tangent space that facilitates the application of regular output layers for the final classification.
Posted Content
Deep Learning on Lie Groups for Skeleton-based Action Recognition
TL;DR: Li et al. as mentioned in this paper incorporated the Lie group structure into a deep network architecture to learn more appropriate Lie group features for skeleton-based action recognition, and designed rotation mapping layers to transform the input Lie group feature into desirable ones, which are aligned better in the temporal domain.
Proceedings ArticleDOI
Dense 3D Regression for Hand Pose Estimation
TL;DR: Zhang et al. as discussed by the authors decompose the pose parameters into a set of per-pixel estimations, i.e., 2D heat maps, 3D heatmaps and unit 3D directional vector fields.
Proceedings ArticleDOI
Crossing Nets: Combining GANs and VAEs with a Shared Latent Space for Hand Pose Estimation
TL;DR: In this article, the authors propose a semi-supervised generative model for 3D hand pose estimation from depth images, where the generator is updated with the back-propagated gradient from the discriminator to synthesize realistic depth maps.
Journal ArticleDOI
MEgATrack: monochrome egocentric articulated hand-tracking for virtual reality
Shangchen Han,Beibei Liu,Randi Cabezas,Christopher D. Twigg,Peizhao Zhang,Jeff Petkau,Tsz-Ho Yu,Chun-Jung Tai,Muzaffer Akbay,Zheng Wang,Asaf Nitzan,Gang Dong,Yuting Ye,Lingling Tao,Chengde Wan,Robert Y. Wang +15 more
TL;DR: This work designs scalable, semi-automated mechanisms to collect a large and diverse set of ground truth data using a combination of manual annotation and automated tracking, and introduces a detection-by-tracking method that increases smoothness while reducing the computational cost.