ICRA 2024 Cloth Competition

Official repository for the ICRA 2024 Cloth Competition 👕.

The dataset can be found here.

For general questions about the competition, please use our Github Discussions page.

Getting started

1. First complete the installation instructions.
2. Then you can run the getting started notebook 📔 to explore the dataset.

Introduction

This repository contains code to help the participants of the ICRA 2024 Cloth Competition prepare for the competition. It contains utilities to load and visualize the dataset, but also all the code we use to actually collect the data. We will also add the code to run the competition and evaluate the performance of the participants.

Some useful starting points:

• 📔 Getting started notebook: notebooks/01_getting_started.ipynb
• 📷 Competition observation definition: cloth-tools/cloth_tools/dataset/format.py
• 👉👕 Manual grasp annotation: notebooks/grasping/01_annotation.ipynb
• 📦 Data collection: scripts/data_collection/09_data_collection_loop.py

Installation

The recommended method to install the required dependencies is to use conda. Make sure you have a version of conda e.g. miniconda installed. To make the conda environment creation faster, we recommend configuring the libmamba solver first.

Then run the following commands:

git clone git@github.com:Victorlouisdg/cloth-competition.git
cd cloth-competition
conda env create -f environment.yaml

To test your installation:

conda activate cloth-competition
jupyter notebook notebooks/01_getting_started.ipynb

Development

Use the environment-dev.yaml file instead, then install the pre-commit hooks to ensure code quality:

pre-commit install

Releasing

Releasing a new version of cloth-tools on PyPi, example:

1. Update the version in pyproject.toml.
2. git tag -a v0.1.0 -m "cloth-tools v0.1.0"
3. git push origin v0.1.0