diff --git a/README.md b/README.md index 53fef9b..92d2341 100644 --- a/README.md +++ b/README.md @@ -5,9 +5,10 @@

-This repository contains the official release of simulation environments and datasets for the [CoRL 2023](https://www.corl2023.org/) paper "MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrations". +This repository contains the official release of simulation environments and datasets for the [CoRL 2023](https://www.corl2023.org/) paper "MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrations". The datasets contain over 48,000 task demonstrations across 12 tasks. -[**[Website]**](https://mimicgen.github.io)   [**[Paper]**](https://openreview.net/forum?id=dk-2R1f_LR) +Website: https://mimicgen.github.io +Paper: https://arxiv.org/abs/2310.17596 ------- @@ -114,6 +115,15 @@ The datasets are split into different types: **Note 2**: We found that the large_interpolation datasets pose a significant challenge for imitation learning, and have substantial room for improvement. +### Dataset Statistics + +We provide more information on the amount of demonstrations for each dataset type: +- **source**: 120 human demonstrations across 12 tasks (10 per task) used to automatically generate the other datasets +- **core**: 26,000 task demonstrations across 12 tasks (26 task variants) +- **object**: 2000 task demonstrations on the Mug Cleanup task with different mugs +- **robot**: 16,000 task demonstrations across 4 different robot arms on 2 tasks (4 task variants) +- **large_interpolation**: 6000 task demonstrations across 6 tasks that pose significant challenges for modern imitation learning methods + ### Dataset Download #### Method 1: Using `download_datasets.py` (Recommended) @@ -142,6 +152,11 @@ You can download the datasets manually through Google Drive. The folders each co **Google Drive folder with all datasets:** [link](https://drive.google.com/drive/folders/14e9kkHGfApuQ709LBEbXrXVI1Lp5Ax7p?usp=drive_link) +#### Method 3: Using Hugging Face + +You can download the datasets through Hugging Face. + +**Hugging Face dataset repository:** [link](https://huggingface.co/datasets/amandlek/mimicgen_datasets) ## Reproducing Policy Learning Results @@ -196,7 +211,7 @@ If you run into an error not documented above, please search through the [GitHub ## Citation -Please cite [the MimicGen paper](https://openreview.net/forum?id=dk-2R1f_LR) if you use this code in your work: +Please cite [the MimicGen paper](https://arxiv.org/abs/2310.17596) if you use this code in your work: ```bibtex @inproceedings{mandlekar2023mimicgen,