README.md

Text-to-Image Person ReID

📚 Contents

• Text-to-Image Person ReID
  • 📚 Contents
  • 📋 Introduction
  • 📂 Datasets
  • 🛠️ Environment
  • 🚀 Get Started
  • 💗 Acknowledgement
  • 🤝 Contribute & Contact

📋 Introduction

The HAP pre-trained model is fine-tuned for the text-to-image person ReID task on CUHK-PEDES, ICFG-PEDES, RSTPReid datasets.

📂 Datasets

Put the dataset directories outside the HAP project:

home
├── HAP
├── CUHK-PEDES  # CUHK-PEDES dataset directory
│   └── imgs
├── ICFG-PEDES  # ICFG-PEDES dataset directory
│   ├── train
│   └── test
└── RSTPReid  # RSTPReid dataset directory
    └── xxx.jpg

🛠️ Environment

Conda is recommended for configuring the environment:

conda env create -f env-text_to_image_person_reid.yaml && conda activate env_t2i_person_reid

🚀 Get Started

We provide BERT checkpoint here.

It may need 1 GPU with memory larger than 14GB, such as NVIDIA V100, for training.

# -------------------- Fine-Tuning HAP for Text-to-Image Person ReID --------------------
cd HAP/downstream/text_to_image_person_reid/

# Download the HAP checkpoint and move it here
# text_to_image_person_reid/ckpt_default_pretrain_pose_mae_vit_base_patch16_LUPersonPose_399.pth

# Download the BERT folder and move it here
# text_to_image_person_reid/bert_base_uncased/

DATASET=CUHK-PEDES  # {CUHK-PEDES, ICFG-PEDES, RSTPReid}
DATA_ROOT=../../../${DATASET}

GPU=0  # Choose an available GPU

python train.py \
  --GPU_id ${GPU} \
  --dataset ${DATASET} \
  --dataroot ${DATA_ROOT}

💗 Acknowledgement

Our implementation is based on the codebase of LGUR.

🤝 Contribute & Contact

Feel free to star and contribute to our repository.

If you have any questions or advice, contact us through GitHub issues or email ([email protected]).