The pretrained model will be released later.
- InterHand
- DualHand
Our repo is based on InterHand2.6M. Below is the related paper list.
# pip switch source
# Tsinghua source
pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
# Ali source
pip config set global.index-url https://mirrors.aliyun.com/pypi/simple/
# Tencent source
pip config set global.index-url http://mirrors.cloud.tencent.com/pypi/simple
# Douban source
pip config set global.index-url http://pypi.douban.com/simple/python create -n hands python=3.8
conda activate hands
# Install Paddle Inference Library
# https://paddle-inference-lib.bj.bcebos.com/2.2.1/python/Windows/GPU/x86-64_vs2017_avx_mkl_cuda11.2_cudnn8/paddlepaddle_gpu-2.2.1.post112-cp38-cp38-win_amd64.whl
pip install paddlepaddle_gpu-2.2.1.post112-cp38-cp38-win_amd64.whl
# Download TesnorRT https://developer.nvidia.com/nvidia-tensorrt-8x-download
# choose TesnorRT8.0
# Installatin guide: https://docs.nvidia.com/deeplearning/tensorrt/install-guide/index.html#installing-zip
# Install other packages
pip install -r requirements.txt
# docker
docker pull paddlepaddle/paddle:2.2.2-gpu-cuda11.2-cudnn8
# create docker environment
# multi GPU training
export CUDA_VISIBLE_DEVICES=0,1,2,3
python -m paddle.distributed.launch --gpus '0,1,2,3' train.py
# single GPU training
python train.py
export ppyolo model.
cd path/to/PaddleDetection
python tools/export_model.py \
--config configs/ppyolo/ppyolo_r18vd_voc.yml \
--opt weights=output/ppyolo_r18vd_voc/best_model.pdparams \
--output_dir path/to/HandInteraction/weightsPretrained Model：None
Download the entire folder from the weight link and put it in the ${model_root_path} / weights folder.
Thanks goes to these wonderful people (emoji key):
 Huan Yang 🚇 |
 Geeksun2018 🚇 |
 bangwhe 💻 🌍 |
This project follows the all-contributors specification. Contributions of any kind welcome!