Add pipeline tag and improve metadata
Browse filesThis PR adds the `image-to-video` pipeline tag to the model card metadata. This ensures the model is correctly categorized on the Hugging Face Hub, making it easier for users to find when filtering by task. I've also added a brief sentence linking to the research paper.
README.md
CHANGED
|
@@ -1,8 +1,10 @@
|
|
| 1 |
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
base_model:
|
| 4 |
- Wan-AI/Wan2.1-T2V-1.3B
|
|
|
|
|
|
|
| 5 |
---
|
|
|
|
| 6 |
<div align="center">
|
| 7 |
<img src="assets/teaser.png">
|
| 8 |
|
|
@@ -25,6 +27,7 @@ Xiaoxu Zheng,
|
|
| 25 |
|
| 26 |
**TL;DR**: 🔭 Single-image long-range view generation via an <u>asynchronous chunk-wise autoregressive diffusion framework</u> that utilizes <u>explicit camera conditioning</u> and <u>online 3D cache</u> for geometric consistency.
|
| 27 |
|
|
|
|
| 28 |
|
| 29 |
## 🎬 Demo Video
|
| 30 |
|
|
|
|
| 1 |
---
|
|
|
|
| 2 |
base_model:
|
| 3 |
- Wan-AI/Wan2.1-T2V-1.3B
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
pipeline_tag: image-to-video
|
| 6 |
---
|
| 7 |
+
|
| 8 |
<div align="center">
|
| 9 |
<img src="assets/teaser.png">
|
| 10 |
|
|
|
|
| 27 |
|
| 28 |
**TL;DR**: 🔭 Single-image long-range view generation via an <u>asynchronous chunk-wise autoregressive diffusion framework</u> that utilizes <u>explicit camera conditioning</u> and <u>online 3D cache</u> for geometric consistency.
|
| 29 |
|
| 30 |
+
This repository contains the weights for **WorldWarp**, presented in [WorldWarp: Propagating 3D Geometry with Asynchronous Video Diffusion](https://arxiv.org/abs/2512.19678).
|
| 31 |
|
| 32 |
## 🎬 Demo Video
|
| 33 |
|