Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
Xingyu Liao fb36b23678 | 3 years ago | |
---|---|---|
.. | ||
configs | 3 years ago | |
naic | 3 years ago | |
README.md | 3 years ago | |
label.txt | 3 years ago | |
naic20r2_train_list_clean.txt | 3 years ago | |
train_list_clean.txt | 3 years ago | |
train_net.py | 3 years ago | |
val_gallery.txt | 3 years ago | |
val_query.txt | 3 years ago |
This repository contains the 1-st place solution of ReID Competition of NAIC. We got the first place in the final stage.
Detailed information about the NAIC competition can be found here.
(2, 1, 0)
, then apply this order for all images of this id, and make it a new id.resnest
is far better than ibn
.resnest101
, resnest200
with different resolution (192x256, 192x384) to ensemble.Before starting, please see GETTING_STARTED.md for the basic setup of FastReID.
All configs are made for 2-GPU training.
python3 projects/NAIC20/train_net.py --config-file projects/NAIC20/configs/r34-ibn.yml --num-gpus 2
MODEL
in submit.yml
toMODEL.WEIGHTS
to the path of your trained model, then run:python3 projects/NAIC20/train_net.py --config-file projects/NAIC20/configs/submit.yml --eval-only --commit --num-gpus 2
You can find submit.json
and distmat.npy
in OUTPUT_DIR
of submit.yml
.
To quickly verify the results, we use resnet34-ibn as backbone to conduct ablation study.
The datasets are naic19
, naic20r1
and naic20r2
.
Setting | Rank-1 | mAP |
---|---|---|
Baseline | 70.11 | 63.29 |
w/ tripletx10 | 73.79 | 67.01 |
w/ cosface | 75.61 | 70.07 |
该算法把样本聚类和特征学习融合到一个端到端的网络框架中,提升模型的跨域能力。该模型在Market训练,DukeMTMC上测试能达到82.0%的准确率,在DukeMTMC上训练,Market上测试能达到92.2%的准确率。
Text Python C++ Protocol Buffer Markdown other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》