FALCON-VLA/CALVIN-3D_cam-params

| FALCON | From Spatial to Actions:
Grounding Vision-Language-Action Model in Spatial Foundation Priors (ICLR 2026)

Zhengshen Zhang   Hao Li   Yalun Dai   Zhengbang Zhu   Lei Zhou  
Chenchen Liu   Dong Wang   Francis E. H. Tay   Sijin Chen  
Ziwei Liu   Yuxiao Liu^*†   Xinghang Li^*   Pan Zhou^*  

^*Corresponding Author  ^†Project Lead

ByteDance Seed
National University of Singapore   Nanyang Technological University
Tsinghua University   Singapore Management University

🚀 Introduction

Existing vision-language-action (VLA) models act in 3D real-world but are typically built on 2D encoders, leaving a spatial reasoning gap that limits generalization and adaptability. In this work, we introduce FALCON (From Spatial to Action), a novel paradigm that injects rich 3D spatial tokens into the action head of a VLA model, enabling robust spatial understanding and SOTA performance across diverse manipulation tasks without disrupting vision-language alignment. See our paper at here.

Dataset Card for CALVIN Dataset-Camera Parameters

Dataset Details

This dataset repo contains preprocessed camera parameters for the static camera under the ABC-D & ABCD-D setting (training/validation) of the Calvin dataset.

Dataset Sources

• Repository: CALVIN Dataset
• Paper: CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Tasks

📦 Usage

We provide an efficient way using the Hugging Face CLI to download the dataset and then extract the final data.

1. Download the dataset

huggingface-cli download FALCON-VLA/CALVIN-3D_cam-params --repo-type dataset

2. Extract the final data

tar -xzvf ./packaged_ABC_D.tar.gz -C ./
tar -xzvf ./packaged_ABCD_D.tar.gz -C ./

After extraction, the dataset files will be available in the current directory.

For camera parameters loading details, please refer to our released data class DiskCalvinDataset_ESM.

Dataset Structure

After extraction, the dataset is organized as follows:

packaged_ABC_D/
├── training/
│   ├── episode_0000001.npz
│   ├── episode_0000002.npz
│   └── ...
└── validation/
    ├── episode_0000038.npz
    ├── episode_0000039.npz
    └── ...

packaged_ABCD_D/
├── training/
│   ├── episode_0000001.npz
│   ├── episode_0000002.npz
│   └── ...
└── validation/
    ├── episode_0000038.npz
    ├── episode_0000039.npz
    └── ...

Each .npz file corresponds to one episode and contains the following fields:

static_cam_ex_mat: static camera extrinsic matrix (camera-to-world), shape: (4, 4)
static_cam_intr_mat: static camera intrinsic matrix, shape: (3, 3)
annotation_id: episode annotation identifier, type: int

🤗 FAQs

If you encounter any issues, feel free to open an issue or reach out through discussions. We appreciate your feedback and contributions! 🚀

🖊️ Citation

If you find this project useful in your research, please consider cite:

@article{zhang2025spatial,
  title={From spatial to actions: Grounding vision-language-action model in spatial foundation priors},
  author={Zhang, Zhengshen and Li, Hao and Dai, Yalun and Zhu, Zhengbang and Zhou, Lei and Liu, Chenchen and Wang, Dong and Tay, Francis EH and Chen, Sijin and Liu, Ziwei and others},
  journal={arXiv preprint arXiv:2510.17439},
  year={2025}
}

@article{mees2022calvin,
  title={Calvin: A benchmark for language-conditioned policy learning for long-horizon robot manipulation tasks},
  author={Mees, Oier and Hermann, Lukas and Rosete-Beas, Erick and Burgard, Wolfram},
  journal={IEEE Robotics and Automation Letters},
  volume={7},
  number={3},
  pages={7327--7334},
  year={2022},
  publisher={IEEE}
}

🪪 License

All datasets, as well as our codebase are released under the Apache-2.0 License.

❤️ Acknowledgement

FALCON is built with reference to the code of the following projects: RoboVLMs, Microsoft Kosmos-2, VGGT, and ManiUniCon. Thanks for their awesome work!