Skip to content

Theory, Experiments, Dataset, and Code for our newly proposed 3D Human Pose, Shape, and Motion reconstruction pipeline

Notifications You must be signed in to change notification settings

HrishavBakulBarua/3D-Human

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

33 Commits
 
 
 
 

Repository files navigation

Enhanced Spatio-Temporal Context for Temporally Consistent Robust 3D Human Motion Recovery from Monocular Videos

This page contains all the information of this project.

About the project

This joint project is a collaboration between TCS Research, Kolkata and IIIT Hyderabad, India.

Project Members -

Sushovan Chanda (TCS Research, Kolkata),
Amogh Tiwari (IIIT Hyderabad),
Lokender Tiwari (TCS Research, Kolkata),
Brojeshwar Bhowmick (TCS Research, Kolkata),
Avinash Sharma (IIIT Hyderabad), and
Hrishav Bakul Barua (TCS Research, Kolkata)&Monash University, Melbourne, Australia

Abstract

Link to the paper!!

Overview

Our method yields superior and temporally consistent motion estimation. It can be observed that our method yield graph (green curve) significantly closer to the ground truth acceleration graph (black curve) compared to existing methods 1, 2, and 3 (result inferred on unseen test video from Human3.6M dataset).

My Image

The Architecture

My Image

Videos of our work

video1.mp4
video2.mp4

Citation

If you find our work in your research or development activities, please cite our paper as follows:


@article{chanda2023enhanced,
  title={Enhanced Spatio-Temporal Context for Temporally Consistent Robust 3D Human Motion Recovery from Monocular Videos},
  author={Chanda, Sushovan and Tiwari, Amogh and Tiwari, Lokender and Bhowmick, Brojeshwar and Sharma, Avinash and Barua, Hrishav},
  journal={arXiv preprint arXiv:2311.11662},
  year={2023}
}

Releases

No releases published

Packages

No packages published