Code for British Machine Vision Conference 2021: Holistic Guidance for Occluded Person ReID
https://www.bmvc2021-virtualconference.com/assets/papers/1087.pdf
#Installation
conda create --name holistic-reid python=3.7
conda activate holistic-reid
# install dependencies
pip install -r requirements.txt
# install torch and torchvision (select the proper cuda version to suit your machine)
conda install pytorch torchvision cudatoolkit=9.0 -c pytorch
# install torchreid (don't need to re-build it if you modify the source code)
python setup.py develop
Steps to reproduce results:
-
Create data directory paths and add data Under root create a directory reid-data Add data Occluded-Duke, Occluded ReID data from https://github.com/lightas/Occluded-DukeMTMC-Dataset, Follow instructions Download market1501 data and follow instructions from https://github.com/lightas/Occluded-DukeMTMC-Dataset
-
Train backbone with source_training.py
-
Use saved model from 2. in the appropriate log folder to train Teacher -Student model using occluded_reid_train.py
Running "occluded_reid_train.py" will train and validate to produce the results required for the given data set. Dataset define in the Key Word "target=" in occluded_reid_train.py