Skip to content

NikeHop/PlaySegmentation-AAAI2025

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🎭✂️ Play Segmentation - AAAI 2025

Codebase for the AAAI-2025 paper "Data Augmentation for Instruction Following Policies via Trajectory Segmentation"

Project Page

Prerequisites

  • Anaconda3/Miniconda3
conda create -n play_segmentation python=3.11
conda activate play_segmentation
pip install -e .

🔍 Overview

Overview of the different components of Play Segmentation

  1. First get the necessary Datasets BabyAI/CALVIN.
  2. Train a Labelling Model on the annotated datasets.
  3. Train Segmentation Models.
  4. Train a policy.

🗂️ Datasets

Calvin

To get the CALVIN dataset follow the instructions here.

BabyAI

To generate the segmented as well as unsegmented trajectories in the BabyAI environment follow the instructions here.

📝 Labelling

Train Labelling Model

Note: The labelling model also needs to be trained to obtain pretrained embeddings of videos for the segmentation methods TriDet and Play Segmentation.

To train the labelling model follow the instructions here.

Label Data

To apply the trained segmentation models to extract labelled segments from play trajectories, follow the instructions here.

✂️ Segmentation

To train the different segmentation models follow the instructions here. At the moment the following segmentation models are implemented:

💪 Policy Training

To train an imitation learning policy for BabyAI and a policy via MCIL for CALVIN see here.

Acknowledgements

Citation

If you use this Codebase, please cite:

@article{hopner2025data,
  title={Data Augmentation for Instruction Following Policies via Trajectory Segmentation},
  author={H{\"o}pner, Niklas and Tiddi, Ilaria and van Hoof, Herke},
  journal={arXiv preprint arXiv:2503.01871},
  year={2025}
}

About

Code for the Paper: Data Augmentation for Instruction Following Policies via Trajectory Segmentation (https://arxiv.org/abs/2503.01871)

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published