Gengshan Yang

high-res-stereo. Hierarchical Deep Stereo Matching on High Resolution Images, CVPR 2019.

rigidmask. Code for "Learning to Segment Rigid Motions from Two Frames". CVPR 2021.

expansion. Upgrading Optical Flow to 3D Scene Flow through Optical Expansion, CVPR 2020 (Oral).

VCN. Volumetric Correspondence Networks for Optical Flow, NeurIPS 2019.

ppr. Code for PPR: Physically Plausible Reconstruction from Monocular Videos (ICCV 23).

rac. RAC: Reconstructing Animatable Categories from Videos. CVPR 2023

viser. ViSER: Video-Specific Surface Embeddings for Articulated 3D Shape Reconstruction. NeurIPS 2021.

monodepth-uncertainty. Inferring distributions over depth from a single image, IROS 2019

ppr-diffphys. This repo contains the differentiable physics simulation module in "PPR: Physically Plausible Reconstruction from Monocular Videos". ICCV 23.

headTracking. Code for detection-based head tracking.

cv-gpu-py. python wrapper for opencv gpu pyramid-lk-flow and tv-l1-flow

dyn_pose. Dynamic pose estimation in MXNet. faster-rcnn detection + cpm pose estimation + LSTM dynamical modeling

UMR. Self-supervised Single-view 3D Reconstruction

kitti-lidar-utils. C++

handProj. Faster-rcnn detection, HOG+SVM tracking and CPM pose estimation. code based on caffe.

Pytorch-Correlation-extension. Custom implementation of Corrleation Module

handShape. hand shape feature extraction from million-hands model

trainSVM. Train an OpenCV-based hog-svm classifier for images.

scRNA-seq-classification. scRNA sequence classification by a 3-layer-MLP. Weighted samples and sigmoid cross entropy loss were used.

caption-eval-general. evaluation of captions for general input