This repo contains the reference source code for the paper Robust Person Re-identification by Modelling Feature Uncertainty in ICCV 2019. In this project, we provide Resnet-Baseline and DistributionNet architectures and results with 10% random noise on Market dataset.
- Python 2.7
- Tensorflow 1.3.0
Download dataset from here
Folder 'Market' includes original Market dataset (training, query, and gallery) and 10% random noise Market training dataset file.
Folder 'result' includes trained models of Resnet-Baseline and DistributionNet
Folder 'pretrained_model' includes Resnet-50 pretrained in ImageNet
Run
bash pipeline.sh
The test results will be recorded in ./result/model_name/rank.txt