Add model card for OmniTryOn

README.md

@@ -0,0 +1,48 @@
+---
+pipeline_tag: image-to-video
+---
+
+# OmniTryOn: Video Try-On Anything at Once!
+
+OmniTryOn is an external-prior-free generative framework designed for the **Try-On Anything** task, which aims to simultaneously transfer diverse wearable objects onto a person in a video in a single inference pass.
+
+- **Paper:** [OmniTryOn: Video Try-On Anything at Once!](https://huggingface.co/papers/2606.08514)
+- **Project Page:** [https://xcltql666.github.io/OmniTryOnProj/](https://xcltql666.github.io/OmniTryOnProj/)
+- **GitHub Repository:** [https://github.com/xcltql666/OminTryOn](https://github.com/xcltql666/OminTryOn)
+- **Dataset:** [TryAny-Bench](https://huggingface.co/datasets/xcll/Try-Any-Bench)
+
+## Overview
+
+OmniTryOn employs a First Frame Wearable Cache strategy to directly provide diverse wearable objects for the generation process through the initial video frame. To maintain consistency, it introduces Spatiotemporally Consistent RoPE (STC-RoPE), establishing spatiotemporal anchors to preserve complex human motions and background dynamics.
+
+## Usage
+
+### Installation
+
+```bash
+conda create -n omnitryon python=3.12
+conda activate omnitryon
+```
+
+### Training
+
+```bash
+bash DiffSynth-Studio\examples\wanvideo\model_training\full\Video-As-Prompt-Wan2.1-14B.sh
+```
+
+### Inference
+
+```bash
+python DiffSynth-Studio\examples\wanvideo\model_training\validate_full\Video-As-Prompt-Wan2.1-14B.py
+```
+
+## Citation
+
+```bibtex
+@article{omnitryon2025,
+  title={OmniTryOn: Video Try-On Anything at Once!},
+  author={},
+  journal={arXiv preprint arXiv:2606.08514},
+  year={2025}
+}
+```