search

workingcoder / MCJA

A New Strong and Simple Baseline Method for VI-ReID (Bridging the Gap: Multi-level Cross-modality Joint Alignment for Visible-infrared Person Re-identification)
MIT License
30 stars 0 forks source link

Unknow file in the main.py #2

Closed elpicio closed 4 months ago

elpicio commented 4 months ago

There is an unknow file in the main.py, line 238" perm = sio.loadmat(os.path.join(cfg.data_root, 'exp', 'rand_perm_cam.mat'))['rand_perm_cam']". The file "rand_perm_cam.mat" seems like in the floder of sysu dataset called 'exp', but I don't find it in the offical dataset of SYSU-MM01. Is this your original code? It seems that I cannot run your code without this file. image

workingcoder commented 4 months ago

This file originates from an official dataset (https://github.com/wuancong/SYSU-MM01/blob/master/evaluation/data_split/rand_perm_cam.mat), and most existing methods utilize this configuration. Thus, we just adopted the same setup for comparison.

elpicio commented 4 months ago

This file originates from an official dataset (https://github.com/wuancong/SYSU-MM01/blob/master/evaluation/data_split/rand_perm_cam.mat), and most existing methods utilize this configuration. Thus, we just adopted the same setup for comparison.

Thank you for your reply. I find that file successfully