Hugging Face's logo

HorizonRobotics
/
RoboTransfer

Image-to-Video
Diffusers
Safetensors
RoboTransferPipeline
Model card Files
xet
 Community
1
nemo04 commited on
Commit
ef6a2e5
·
verified ·
1 Parent(s): 4f510f3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -10
README.md CHANGED
@@ -36,12 +36,14 @@ library_name: diffusers
36
  </div>
37
 
38
  <div align="center">
39
- <img src="assets/pin.png" width="90%" alt="RoboTransfer"/></div>
40
 
41
  ---
42
 
43
  ## 🔍 Abstract
44
 
 
 
45
  **RoboTransfer** is a novel diffusion-based video generation framework tailored for robotic visual policy transfer. Unlike conventional approaches, RoboTransfer introduces **geometry-aware synthesis** by injecting **depth and normal priors**, ensuring multi-view consistency across dynamic robotic scenes. The method further supports **explicit control over scene components**, such as **background editing**, **object identity swapping**, and **motion specification**, offering a fine-grained video generation pipeline that benefits embodied learning.
46
 
47
  ---
@@ -55,15 +57,6 @@ library_name: diffusers
55
 
56
  ---
57
 
58
-
59
- ## 📸 Framework Overview
60
-
61
- ![RoboTransfer Pipeline](assets/robotransfer_pipeline.png)
62
-
63
- > The overall architecture includes view-specific encoding, geometry injection, diffusion denoising with spatial constraints, and component-level editing modules. Our system enables compositional control over scene dynamics while preserving physical and geometric consistency.
64
-
65
- ---
66
-
67
  ## 📖 BibTeX
68
 
69
  ```bibtex
 
36
  </div>
37
 
38
  <div align="center">
39
+ <img src="assets/pin.jpeg" width="50%" alt="RoboTransfer"/></div>
40
 
41
  ---
42
 
43
  ## 🔍 Abstract
44
 
45
+ ![RoboTransfer Pipeline](assets/robotransfer_pipeline.jpeg)
46
+
47
  **RoboTransfer** is a novel diffusion-based video generation framework tailored for robotic visual policy transfer. Unlike conventional approaches, RoboTransfer introduces **geometry-aware synthesis** by injecting **depth and normal priors**, ensuring multi-view consistency across dynamic robotic scenes. The method further supports **explicit control over scene components**, such as **background editing**, **object identity swapping**, and **motion specification**, offering a fine-grained video generation pipeline that benefits embodied learning.
48
 
49
  ---
 
57
 
58
  ---
59
 
 
 
 
 
 
 
 
 
 
60
  ## 📖 BibTeX
61
 
62
  ```bibtex