gengshan-y Goto Github PK

3d-coded

Pytorch Implementation for the project : `3D-CODED` and `Learning Elementary Structure`

adaptive-huffman-coding

aho-corasick-automaton

av-gnn

banmo

BANMo Building Animatable 3D Neural Models from Many Casual Videos

caption-eval-general

evaluation of captions for general input

captionpruning

provides useful tool to process the subtitles and time stamps of a video data set.

cmu-vision.github.io

cv-gpu-py

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

droid-slam

dyn_pose

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

expansion

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

gsplat-dev

handproj

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

handshape

hand shape feature extraction from million-hands model

headtracking

Code for detection-based head tracking.

hgposeextr

Modified hourglass. Extract pose estimations.

high-res-stereo

Hierarchical Deep Stereo Matching on High Resolution Images, CVPR 2019.

hmr2-dev

homework

Imitation Learning Homework 1

kitti-lidar-utils

minvis

monodepth

Unsupervised single image depth prediction with CNNs

monodepth-uncertainty

Inferring distributions over depth from a single image, IROS 2019

monodepth2

Monocular depth estimation from a single image

nerfies

This is the code for Deformable Neural Radiance Fields, a.k.a. Nerfies.

neuraltalk2_generalinput

This is a simplified version of Neuraltalk2. It takes general-features/sentence pairs as input, and uses them to train a language model.

os-lab

posecaptioner_train

Train sign language captioner with pose features.

ppr

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