Home

Awesome

A2J-Transformer

Introduction

This is the official implementation for the paper, "A2J-Transformer: Anchor-to-Joint Transformer Network for 3D Interacting Hand Pose Estimation from a Single RGB Image", CVPR 2023.

Paper link here: A2J-Transformer: Anchor-to-Joint Transformer Network for 3D Interacting Hand Pose Estimation from a Single RGB Image

About our code

Updates

Installation and Setup

Requirements

Compiling CUDA operators(Following Deformable-DETR)

cd ./dab_deformable_detr/ops
sh make.sh

Usage

Dataset preparation

Testing on InterHand 2.6M Dataset

NYU and HANDS 2017 dataset

Cite

Our code is protected by patents and cannot be used for commercial purposes. If you have commercial needs, please contact Prof. Yang Xiao (Huazhong University of Science and Technology): Yang_Xiao@hust.edu.cn.

If you find our work useful in your research or publication, please cite our work:

@inproceedings{jiang2023a2j,
  title={A2J-Transformer: Anchor-to-Joint Transformer Network for 3D Interacting Hand Pose Estimation from a Single RGB Image},
  author={Jiang, Changlong and Xiao, Yang and Wu, Cunlin and Zhang, Mingyang and Zheng, Jinghong and Cao, Zhiguo and Zhou, Joey Tianyi},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={8846--8855},
  year={2023}
}