Add model card for DADP

README.md

@@ -0,0 +1,49 @@
+---
+pipeline_tag: reinforcement-learning
+tags:
+- robotics
+- control
+- diffusion-policy
+---
+
+# DADP: Domain Adaptive Diffusion Policy
+
+[DADP](https://huggingface.co/papers/2602.04037) is a framework for learning domain-adaptive policies that can generalize to unseen transition dynamics in learning-based control. It achieves robust adaptation through unsupervised disentanglement and domain-aware diffusion injection.
+
+- **Paper:** [DADP: Domain Adaptive Diffusion Policy](https://huggingface.co/papers/2602.04037)
+- **Project Page:** [https://outsider86.github.io/DomainAdaptiveDiffusionPolicy/](https://outsider86.github.io/DomainAdaptiveDiffusionPolicy/)
+- **Repository:** [https://github.com/QinghangLiu/DADP_official](https://github.com/QinghangLiu/DADP_official)
+
+## Overview
+
+DADP introduces two key innovations to tackle the challenge of zero-shot adaptation:
+1. **Lagged Context Dynamical Prediction**: A strategy that conditions future state estimation on a historical offset context to unsupervisedly disentangle static domain representations by filtering out transient properties.
+2. **Domain-Aware Diffusion Injection**: Integration of the learned domain representations directly into the generative process by biasing the prior distribution and reformulating the diffusion target.
+
+The framework has been evaluated on challenging benchmarks across locomotion (Ant, Walker2d, HalfCheetah, Hopper) and manipulation (Door, Relocate).
+
+## Usage
+
+For detailed instructions on environment setup, downloading datasets, and running training or evaluation scripts, please refer to the [official GitHub repository](https://github.com/QinghangLiu/DADP_official).
+
+### Supported Environments
+The training and extraction scripts support the following `env_key` variables:
+- `ant`
+- `halfcheetah`
+- `walker`
+- `hopper`
+- `door`
+- `relocate`
+
+## Citation
+
+If you find this work useful, please consider citing:
+
+```bibtex
+@article{liu2025dadp,
+  title={DADP: Domain Adaptive Diffusion Policy},
+  author={Liu, Qinghang and others},
+  journal={arXiv preprint arXiv:2602.04037},
+  year={2025}
+}
+```