leejeongho3214 / synthetic-hand-dataset-including-a-forearm

Enhancing 3D Hand Pose Estimation Using SHaF: Synthetic Hand Dataset Including a Forearm (APIN 2023 under review...) ✨✨

It is a novel synthetic hand dataset called Synthetic Hand Dataset Including a Forearm (SHaF) and an efficient transformer-based three-dimensional (3D) hand pose estimation model. This model achieves the superior performance of the estimation model over that of previous state-of-the-art (SOTA) models.

Directory

Build as the below architecture

{$ROOT}
|-- build
|-- src
|-- datasets
|-- models
|-- docs

Setup with Conda

git clone https://github.com/leejeongho3214/Wearable_Pose_Model.git
cd Wearable_Pose_Model
conda env create -f requirements.yaml

Model Download

Please download our files that need to run our code. download

Train

If you locates in Wearabel_Pose_Model folder, run the code below and the training will begin

cd src/tools
python train.py

Solution

If you get a error message such as the wrong path, follow the code below

import sys
sys.path.append("/usr/your/path/Wearable_Pose_Model")

Result

Evaluation Metric

• PA-MPJPE
• AUC

Quantitative Evaluation