search

michuanhaohao / AlignedReID

Alignedreid＋+: Dynamically Matching Local Information for Person Re-Identification.
MIT License
401 stars 114 forks source link

Hello? I have some doubts about my results on CUHK03 #14

Closed romanticegg closed 5 years ago

romanticegg commented 5 years ago

I have done some improvements on your network structure, but the results on CUHK03(detected) have surprised me, it got 84.6%mAP and 77.3% rank 1 accuracy for CUHK03. I don't believe this result so I doubt whether there is anything wrong with my command? My command is as follows: python train_alignedreid.py -d cuhk03 --use-metric-cuhk03 -a resnet50 --test_distance global_local --stepsize 300 --lr 0.0002 --save-dir ./log/try/ --train-batch 128 --eval-step 150 --max-epoch 60000 --reranking --height 384 --width 128

michuanhaohao commented 5 years ago

@romanticegg If you want compare the results with some published papers on CUHK03, you should remove ' --use-metric-cuhk03'. You can get more details from the 'Readme.md' file.

Old test protocol: ' --use-metric-cuhk03' + '--cuhk03-classic-split' New test protocol: default (do nothing)

romanticegg commented 5 years ago

@michuanhaohao Oh I know the problem. You are nice! Thanks a lot!