Code for the IJCAI 2018 paper Cross-Modality Person Re-Identification with Generative Adversarial Training . This model performs adversarial learning on multi-modal images (RGB and Infrared) to perform re-identification across modalities.
This code is under active development (after almost 2 years). I hope to add results and models by summer 2021! Stay tuned :)