The code will be released. This repository is Pytorch code for our proposed Baseline method for Cross-Modality Package Re-Identification.
The dataset RX01 is privacy sensitive and is currently only supported for academic research. If you need it, please contact us. [email protected].
Train a model by
python train.py
--dataset
: which dataset "RX01" or "sysu".
The results may have some fluctuation, and might be better by finetuning the hyper-parameters.
Datasets | Rank@1 | mAP |
---|---|---|
#RX01 | ~ 48.0% | ~ 59.0% |
#SYSU-MM01[1] | ~ 68.54% | ~ 66.53% |
Our code is based on Yukang Zhang [2].
[1] A. Wu, W.-s. Zheng, H.-X. Yu, S. Gong, and J. Lai. Rgb-infrared crossmodality person re-identification. In IEEE International Conference on Computer Vision (ICCV), pages 5380โ5389, 2017.
[2] Zhang, Yukang, et al. "Towards a unified middle modality learning for visible-infrared person re-identification." Proceedings of the 29th ACM International Conference on Multimedia. 2021.