search

jiawei151 / SGIEL_VIReID

Official PyTorch Implementation of "Shape-Erased Feature Learning for Visible-Infrared Person Re-Identification" (CVPR'23)
MIT License
38 stars 3 forks source link

readme

SGIEL_VIReID (CVPR 2023)

Official PyTorch Implementation of "Shape-Erased Feature Learning for Visible-Infrared Person Re-Identification" (CVPR'23)

Datasets

We follow Cross-Modal-Re-ID-baseline to preprocess SYSU-MM01 dataset.

For VCM-HITSZ, please refer to its official repository.

Body Shape Data

We borrowed pre-trained Self-Correction Human Parsing (SCHP) model (pretrained on Pascal-Person-Part dataset) to segment body shape from background. Given a pixel of a visible or infrared image, we directly summed the probabilities of being a part of the head, torso, or limbs, predicted by SCHP, to create the body-shape map.

You can also download the body shape data for SYSU-MM01 through this link.

Dependencies

Training and Model

To reproduce our results on SYSU-MM01, just run (after the dataset path declared)

bash run.sh

We are currently working on Issues. Please feel free to contact me (fengjw151@gmail.com) if you need any other information.

We uploaded a trained model on SYSU-MM01.

Acknowledge

Thanks for the great code base from the open-sourced Cross-Modal-Re-ID-baseline.