Dataset Viewer
The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.
Couldn't infer the same data file format for all splits. Got {NamedSplit('train'): ('parquet', {}), NamedSplit('validation'): ('imagefolder', {})}
Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

IROS-2025-Challenge-Manip

Dataset Summary πŸ“–

This dataset contains the IROS Challenge - Manipulation Track benchmark, organized into pretrain, train, and validation splits.

  • Pretrain split: ~20,000 single pick-and-place trajectories, packaged into tar files (each containing ~1,000 trajectories).
  • Train split: task-specific demonstrations, with ~100 trajectories provided per task.
  • Validation split: includes the test-time scenes and object assets in USD format.

Each trajectory in the pretrain and train splits contains:

  • Multi-view video recordings (three perspectives: head-mounted camera and two wrist cameras)
  • Robot states (joint positions, gripper states, etc.)
  • Actions corresponding to the task execution

This dataset is designed to support pretraining, task-specific fine-tuning, and evaluation for robotic manipulation in the IROS Challenge setting.

Get started πŸ”₯

Download the Dataset

To download the full dataset, you can use the following code. If you encounter any issues, please refer to the official Hugging Face documentation.

from huggingface_hub import snapshot_download

dataset_path = snapshot_download("InternRobotics/IROS-2025-Challenge-Manip", repo_type="dataset")

Please execute this Python file to post-process the validation set.

cd IROS-2025-Challenge-Manip
python dataset_post_processing.py validation

Unzip the pretrain dataset

cd pretrain
for i in {1..20}; do
    echo "Extracting $i.tar.gz ..."
    tar -xzf "$i.tar.gz"
done

Dataset Structure

pretrain Folder hierarchy

pretrain
β”œβ”€β”€ 1.tar.gz
β”‚   └── 1/
β”‚       β”œβ”€β”€ data/
β”‚       β”œβ”€β”€ meta/
β”‚       └── videos/
β”œβ”€β”€ 2.tar.gz
β”‚   └── 2/
β”‚       β”œβ”€β”€ data/
β”‚       β”œβ”€β”€ meta/
β”‚       └── videos/
...
β”œβ”€β”€ 20.tar.gz
    └── 20/
        β”œβ”€β”€ data/
        β”œβ”€β”€ meta/
        └── videos/

train Folder hierarchy

train
β”œβ”€β”€ collect_three_glues
β”‚   β”œβ”€β”€ data/
β”‚   β”œβ”€β”€ meta/
β”‚   └── videos/
β”œβ”€β”€ collect_two_alarm_clocks/
β”œβ”€β”€ collect_two_shoes/
β”œβ”€β”€ gather_three_teaboxes/
β”œβ”€β”€ make_sandwich/
β”œβ”€β”€ oil_painting_recognition/
β”œβ”€β”€ organize_colorful_cups/
β”œβ”€β”€ purchase_gift_box/
β”œβ”€β”€ put_drink_on_basket/
└── sort_waste/

validation Folder hierarchy

validation
β”œβ”€β”€ IROS_C_V3_Aloha_seen
β”‚   β”œβ”€β”€ collect_three_glues
β”‚   β”‚   β”œβ”€β”€ 000
β”‚   β”‚   β”‚   β”œβ”€β”€ meta_info.pkl
β”‚   β”‚   β”‚   β”œβ”€β”€ scene.usd
β”‚   β”‚   β”‚   └── SubUSDs -> ../SubUSDs
β”‚   β”‚   β”œβ”€β”€ 001/
β”‚   β”‚   β”œβ”€β”€ 002/
β”‚   β”‚   β”œβ”€β”€ 003/
β”‚   β”‚   β”œβ”€β”€ 004/
β”‚   β”‚   β”œβ”€β”€ 005/
β”‚   β”‚   β”œβ”€β”€ 006/
β”‚   β”‚   β”œβ”€β”€ 007/
β”‚   β”‚   β”œβ”€β”€ 008/
β”‚   β”‚   β”œβ”€β”€ 009/
β”‚   β”‚   └── SubUSDs
β”‚   β”‚       β”œβ”€β”€ materials/
β”‚   β”‚       └── textures/
β”‚   β”œβ”€β”€ collect_two_alarm_clocks/
β”‚   β”œβ”€β”€ collect_two_shoes/
β”‚   β”œβ”€β”€ gather_three_teaboxes/
β”‚   β”œβ”€β”€ make_sandwich/
β”‚   β”œβ”€β”€ oil_painting_recognition/
β”‚   β”œβ”€β”€ organize_colorful_cups/
β”‚   β”œβ”€β”€ purchase_gift_box/
β”‚   β”œβ”€β”€ put_drink_on_basket/
β”‚   └── sort_waste/
└── IROS_C_V3_Aloha_unseen
    β”œβ”€β”€ collect_three_glues/
    β”œβ”€β”€ collect_two_alarm_clocks/
    β”œβ”€β”€ collect_two_shoes/
    β”œβ”€β”€ gather_three_teaboxes/
    β”œβ”€β”€ make_sandwich/
    β”œβ”€β”€ oil_painting_recognition/
    β”œβ”€β”€ organize_colorful_cups/
    β”œβ”€β”€ purchase_gift_box/
    β”œβ”€β”€ put_drink_on_basket/
    └── sort_waste/

License and Citation

All the data and code within this repo are under CC BY-NC-SA 4.0. Please consider citing our project if it helps your research.

@misc{contributors2025internroboticsrepo,
  title={IROS-2025-Challenge-Manip Colosseum},
  author={IROS-2025-Challenge-Manip Colosseum contributors},
  howpublished={\url{https://github.com/internrobotics/IROS-2025-Challenge-Manip}},
  year={2025}
}
Downloads last month
8,423

Collection including InternRobotics/IROS-2025-Challenge-Manip