1. ubuntu installs anaconda3 5.X
2. build virtual envs
> conda create -n bagreid python=3.6
> source activate bagreid
3. conda install package
> conda install pytorch torchvision cudatoolkit=9.0 -c pytorch
> conda install pandas
> pip install tensorboardX yacs
4. > git clone https://github.com/NVIDIA/apex
> cd apex
> pip install -v --no-cache-dir ./
1. download mvb dataset from http://volumenet.cn/#/
2. > git clone https://github.com/wuyuejinxia/prcv2019-mvb-renet.git
> cd prcv2019-mvb-renet
> mkdir data
> mv MVB_train/Image data/MVB/bounding_box_train
> mv MVB_train/Info/train.json data/MVB/train.json
> mv MVB_val/Image/gallery data/MVB/gallery
> mv MVB_val/Info/val_gallery.json data/MVB/val_gallery.json
> mv MVB_val/Image/probe data/MVB/probe
> mv MVB_val/Info/val_probe.json data/MVB/val_probe.json
> python train.py
| Rank-1 | Rank-3 | Rank-5 | Rank-10 | |
|---|---|---|---|---|
| 2019-07-27 | 84.32% | 94.20% | 96.67% | 98.76% |
| 2019-08-20 | 84.98% | 95.44% | 97.24% | 98.86% |
| 2019-08-29 | 88.69% | 96.20% | 98.38% | 99.05% |
You are encouraged to cite the following papers if this work helps your research.
@misc{zhang2019mvb,
title={MVB: A Large-Scale Dataset for Baggage Re-Identification and Merged Siamese Networks},
author={Zhulin Zhang and Dong Li and Jinhua Wu and Yunda Sun and Li Zhang},
year={2019},
eprint={1907.11366},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
url: https://arxiv.org/abs/1907.11366