Not so strong baseline for Video-Person-ReID

Introduction

This repository contains a not so strong baseline for video-based person reID. It is mainly forked from video-person-reid and reid-strong-baseline. What i done is just merging them ;) and introducing the nvidia-apex to convert the model into a syncbn_model as well as doing slight modifies on the model and tricks, the reason of introducing apex is mainly because my poverty, if you owes a 32GB V100 Graphic Card, you can just ignore the apex operation and run the codes on a single card, then i nearly contributes nothing in this work :).

Requirements:

pytorch >= 0.4.1
torchvision >= 0.2.1
tqdm
[nvidia-apex](https://github.com/NVIDIA/apex), please follow the detailed install instructions 

Dataset

MARS

Experiments on MARS, as it is the largest dataset available to date for video-based person reID. Please follow deep-person-reid to prepare the data. The instructions are copied here:

1. Create a directory named mars/.
2. Download dataset to mars/ from http://www.liangzheng.com.cn/Project/project_mars.html.
3. Extract bbox_train.zip and bbox_test.zip.
4. Download split information from https://github.com/liangzheng06/MARS-evaluation/tree/master/info and put info/ in data/mars (we want to follow the standard split in [8]). The data structure would look like:

   mars/
       bbox_test/
       bbox_train/
       info/
   

5. Change the global variable _C.DATASETS.ROOT_DIR to /path2mars/mars and _C.DATASETS.NAME to mars in config or configs.

Duke-VID

1. Create a directory named duke/ under data/.
2. Download dataset to data/duke/ from http://vision.cs.duke.edu/DukeMTMC/data/misc/DukeMTMC-VideoReID.zip.
3. Extract DukeMTMC-VideoReID.zip.

   duke/
       train/
       gallery/
       query/
   

4. Change the global variable _C.DATASETS.ROOT_DIR to /path2duke/duke and _C.DATASETS.NAME to duke in config or configs.

Usage

To train the model, please run(TP)

python main_baseline.py

To train the model, please run(TA)

python main_baseline.py --at

To train the model, please run(TP + IFFR)

python main_baseline.py --iffr

To train the model, please run(TA + IFFR)

python main_baseline.py --at --iffr

If you want to use surf features to perform cluster filtering of key frames, please add

--surf

Please modifies the settings directly on the config files.
(to be complete)

Performance

Best performance on MARS: mAP : 81.2% Rank-1 : 86.6% Rank-5 : 96.0%

Ablation study and experiments on Duke-VID is undergoing,

Since I'm graduated, I have no graphic cards to conduct the expriments, my sincere aplogies....Hoping the merge request from U, my rich friends~