Add model card for AnyPos

README.md

@@ -0,0 +1,53 @@
+---
+pipeline_tag: robotics
+---
+
+# AnyPos: Automated Task-Agnostic Actions for Bimanual Manipulation
+
+AnyPos is a unified pipeline that integrates large-scale automated task-agnostic exploration with robust embodiment modeling through inverse dynamics learning. It learns embodiment dynamics directly from task-agnostic action data and decouples them from high-level policy learning, enabling scalable reuse of action data across different tasks and platforms.
+
+- **Paper:** [AnyPos: Automated Task-Agnostic Actions for Bimanual Manipulation](https://huggingface.co/papers/2507.12768)
+- **Project Page:** [https://embodiedfoundation.github.io/vidar_anypos](https://embodiedfoundation.github.io/vidar_anypos)
+- **Repository:** [https://github.com/yaofeng1998/Vidar](https://github.com/yaofeng1998/Vidar)
+
+## Introduction
+
+Learning generalizable manipulation policies hinges on data, yet robot manipulation data is scarce and often entangled with specific embodiments. AnyPos tackles this challenge by focusing on exploring all feasible actions of the embodiment to capture what is physically feasible and consistent. Building on this principle, AnyPos generates diverse yet safe trajectories at scale, then learns embodiment representations by decoupling arm and end-effector motions. The pipeline can be seamlessly coupled with diverse high-level policy models.
+
+## Environment Setup
+
+To set up the environment for Vidar/AnyPos, follow these steps:
+
+```bash
+conda create -n vidar python==3.11.9
+conda activate vidar
+conda install pytorch==2.4.0 torchvision==0.19.0 torchaudio==2.4.0 pytorch-cuda=12.4 -c pytorch -c nvidia
+python -m pip install -r requirements.txt
+```
+
+## Usage
+
+The repository provides scripts for inference and evaluation of the Video Diffusion Model (VM) and the Masked Inverse Dynamic Model (IDM).
+
+### Video Diffusion Model Inference
+To generate a video based on an initial frame and instruction:
+```bash
+scripts/vm/sample.sh
+```
+
+### Masked Inverse Dynamic Model Evaluation
+To evaluate the model:
+```bash
+scripts/idm/eval.sh
+```
+
+## Citation
+
+```bibtex
+@article{zhou2025anypos,
+  title={AnyPos: Automated Task-Agnostic Actions for Bimanual Manipulation},
+  author={Zhou, Feng and others},
+  journal={arXiv preprint arXiv:2507.12768},
+  year={2025}
+}
+```