Text-Based Person Retrieval via Cross-Modality Redundant Perception

Usage

Requirements

we use single RTX3090 24G GPU for training and evaluation.

pytorch 1.9.0
torchvision 0.10.0
prettytable
easydict

Prepare Datasets

Download the CUHK-PEDES dataset from here, ICFG-PEDES dataset from here and RSTPReid dataset form here

Organize them in your dataset root dir folder as follows:

|-- your dataset root dir/
|   |-- <CUHK-PEDES>/
|       |-- imgs
|            |-- cam_a
|            |-- cam_b
|            |-- ...
|       |-- reid_raw.json
|
|   |-- <ICFG-PEDES>/
|       |-- imgs
|            |-- test
|            |-- train 
|       |-- ICFG_PEDES.json
|
|   |-- <RSTPReid>/
|       |-- imgs
|       |-- data_captions.json

Training

python train.py

Testing

python test.py

Comparison with other methods on three datasets (CUHK-PEDES, ICFG-PEDES, and RSTPReid). Rank-1, Rank-5, and Rank-10 represent the accuracy (%), with higher values indicating better performance.

Model & log for CUHK-PEDES

Model & log for ICFG-PEDES

Model & log for RSTPReid

Acknowledgments

Some components of this code implementation are adopted from CLIP, IRRA and TransReID. We sincerely appreciate for their contributions.