Skip to content
Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation
Python Dockerfile
Branch: master
Clone or download
Latest commit 516ee50 Jul 13, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
advent release ADVENT Jul 13, 2019
.gitignore Initial commit Apr 4, 2019
Dockerfile release ADVENT Jul 12, 2019
LICENSE release ADVENT Jul 12, 2019
README.md release ADVENT Jul 12, 2019
setup.py release ADVENT Jul 12, 2019
teaser.jpg release ADVENT Jul 12, 2019
tox.ini release ADVENT Jul 12, 2019

README.md

ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation

Paper

ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation
Tuan-Hung Vu, Himalaya Jain, Maxime Bucher, Matthieu Cord, Patrick Pérez
valeo.ai, France
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019 (Oral)

If you find this code useful for your research, please cite our paper:

@inproceedings{vu2018advent,
  title={ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation},
  author={Vu, Tuan-Hung and Jain, Himalaya and Bucher, Maxime and Cord, Mathieu and P{\'e}rez, Patrick},
  booktitle={CVPR},
  year={2019}
}

Abstract

Semantic segmentation is a key problem for many computer vision tasks. While approaches based on convolutional neural networks constantly break new records on different benchmarks, generalizing well to diverse testing environments remains a major challenge. In numerous real world applications, there is indeed a large gap between data distributions in train and test domains, which results in severe performance loss at run-time. In this work, we address the task of unsupervised domain adaptation in semantic segmentation with losses based on the entropy of the pixel-wise predictions. To this end, we propose two novel, complementary methods using (i) an entropy loss and (ii) an adversarial loss respectively. We demonstrate state-of-the-art performance in semantic segmentation on two challenging synthetic-2-real set-ups and show that the approach can also be used for detection.

Demo

Preparation

Pre-requisites

  • Python 3.7
  • Pytorch >= 0.4.1
  • CUDA 9.0 or higher

Installation

  1. Clone the repo:
$ git clone https://github.com/valeoai/ADVENT
$ cd ADVENT
  1. Install OpenCV if you don't already have it:
$ conda install -c menpo opencv
  1. Install this repository and the dependencies using pip:
$ pip install -e <root_dir>

With this, you can edit the ADVENT code on the fly and import function and classes of ADVENT in other project as well.

  1. Optional. To uninstall this package, run:
$ pip uninstall ADVENT

You can take a look at the Dockerfile if you are uncertain about steps to install this project.

Datasets

By default, the datasets are put in <root_dir>/data. We use symlinks to hook the ADVENT codebase to the datasets. An alternative option is to explicitlly specify the parameters DATA_DIRECTORY_SOURCE and DATA_DIRECTORY_TARGET in YML configuration files.

  • GTA5: Please follow the instructions here to download images and semantic segmentation annotations. The GTA5 dataset directory should have this basic structure:
<root_dir>/data/GTA5/                               % GTA dataset root
<root_dir>/data/GTA5/images/                        % GTA images
<root_dir>/data/GTA5/labels/                        % Semantic segmentation labels
...
  • Cityscapes: Please follow the instructions in Cityscape to download the images and validation ground-truths. The Cityscapes dataset directory should have this basic structure:
<root_dir>/data/Cityscapes/                         % Cityscapes dataset root
<root_dir>/data/Cityscapes/leftImg8bit              % Cityscapes images
<root_dir>/data/Cityscapes/leftImg8bit/val
<root_dir>/data/Cityscapes/gtFine                   % Semantic segmentation labels
<root_dir>/data/Cityscapes/gtFine/val
...

Pre-trained models

Pre-trained models can be downloaded here and put in <root_dir>/pretrained_models

Running the code

For evaluation, execute:

$ cd <root_dir>/advent/scripts
$ python test.py --cfg ./configs/advent_pretrained.yml
$ python test.py --cfg ./configs/minent_pretrained.yml
$ python test.py --cfg ./configs/advent+minent.yml

Training

For the experiments done in the paper, we used pytorch 0.4.1 and CUDA 9.0. To ensure reproduction, the random seed has been fixed in the code. Still, you may need to train a few times to reach the comparable performance.

By default, logs and snapshots are stored in <root_dir>/experiments with this structure:

<root_dir>/experiments/logs
<root_dir>/experiments/snapshots

To train AdvEnt:

$ cd <root_dir>/advent/scripts
$ python train.py --cfg ./configs/advent.yml
$ python train.py --cfg ./configs/advent.yml --tensorboard         % using tensorboard

To train MinEnt:

$ python train.py --cfg ./configs/minent.yml
$ python train.py --cfg ./configs/minent.yml --tensorboard         % using tensorboard

Testing

To test AdvEnt:

$ cd <root_dir>/advent/scripts
$ python test.py --cfg ./configs/advent.yml

To test MinEnt:

$ python test.py --cfg ./configs/minent.yml

Acknowledgements

This codebase is heavily borrowed from AdaptSegNet and Pytorch-Deeplab.

License

ADVENT is released under the Apache 2.0 license.

You can’t perform that action at this time.