https://github.com/compvis/unsupervised-part-segmentation
Code for GCPR 2020 Oral : "Unsupervised Part Discovery by Unsupervised Disentanglement"
https://github.com/compvis/unsupervised-part-segmentation
birds computer-vision cub gcpr humans image-segmentation machine-learning paper pytorch tensorflow unsupervised
Last synced: 10 months ago
JSON representation
Code for GCPR 2020 Oral : "Unsupervised Part Discovery by Unsupervised Disentanglement"
- Host: GitHub
- URL: https://github.com/compvis/unsupervised-part-segmentation
- Owner: CompVis
- Created: 2020-09-09T12:55:34.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2020-10-02T06:28:20.000Z (over 5 years ago)
- Last Synced: 2025-03-21T12:07:18.356Z (11 months ago)
- Topics: birds, computer-vision, cub, gcpr, humans, image-segmentation, machine-learning, paper, pytorch, tensorflow, unsupervised
- Language: Jupyter Notebook
- Homepage: https://compvis.github.io/unsupervised-part-segmentation/
- Size: 52.6 MB
- Stars: 34
- Watchers: 6
- Forks: 4
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Unsupervised Part Discovery by Unsupervised Disentanglement
Code accompanying the [GCPR 2020]() paper
[**Unsupervised Part Discovery by Unsupervised Disentanglement**](https://compvis.github.io/unsupervised-part-segmentation/index.html)
[Sandro Braun](https://github.com/therealsupermario),
[Patrick Esser](https://github.com/pesser),
[Björn Ommer](https://hci.iwr.uni-heidelberg.de/Staff/bommer)
[arXiv](https://arxiv.org/abs/2009.04264) | [BibTeX](#bibtex) | [Project Page](https://compvis.github.io/unsupervised-part-segmentation/index.html)
Table of Contents
=================
* [Requirements](#requirements)
* [Training](#training)
* [Data](#data)
* [Evaluation](#evaluation)
* [Pretrained Models](#pretrained-models)
* [BibTeX](#bibtex)
## Requirements
A suitable [conda](https://conda.io/) environment named `braun20parts` can be created
and activated with:
```bash
conda env create -f environment.yaml
conda activate braun20parts
```
Clone the repo **with all it's submodules**
```bash
git clone --recursive -j8 git@github.com:CompVis/unsupervised-part-segmentation.git
```
## Training
1. For running experiments into the respective subfolders `deepfashion`, `cub` and `pennaction`.
2. Experiments can be run using [edflow](https://github.com/edflow/edflow).
```
edflow -t xxx/
```
### Data
* The CUB dataset is the same as in [Lorenz19](https://arxiv.org/pdf/1903.06946.pdf), but we manually added semantic part segmentations and added them in the repo.
## Evaluation
1. baseline models with pretrained checkpoints on all datasets can be found in folder `baselines`
2. evaluation scripts and notebooks can be found in folder `evaluation`
## Pretrained Models
pretrained models can be found in the respective folder, under `train/checkpoints`
## BibTex
```
@inproceedings{braun2020parts,
title={Unsupervised Part Discovery by Unsupervised Disentanglement},
author={Braun, Sandro and Esser, Patrick and Ommer, Bj{\"o}rn},
booktitle={Proceedings of the German Conference on Computer Vision},
year={2020}
}
```