tiangexiang/OccFusion
[NeurIPS 2024] OccFusion: Rendering Occluded Humans with Generative Diffusion Priors
This project helps visual effects artists or content creators reconstruct complete, realistic 3D human models from video footage where parts of the person are hidden or blocked. You input a monocular video showing a person with occlusions, and it outputs a high-fidelity 3D rendering of the fully visible human, suitable for animations or digital scenes. It's designed for professionals working with digital humans in creative fields.
No commits in the last 6 months.
Use this if you need to create realistic 3D digital doubles or animations of people from existing video footage, even when parts of them are obscured.
Not ideal if you're looking for a general-purpose object segmentation or 2D image inpainting tool rather than 3D human reconstruction.
Stars
18
Forks
4
Language
Python
License
—
Category
Last pushed
Nov 24, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/tiangexiang/OccFusion"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
PRIS-CV/DemoFusion
Let us democratise high-resolution generation! (CVPR 2024)
mit-han-lab/distrifuser
[CVPR 2024 Highlight] DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Models
Tencent-Hunyuan/HunyuanPortrait
[CVPR-2025] The official code of HunyuanPortrait: Implicit Condition Control for Enhanced...
giuvecchio/matfuse
MatFuse: Controllable Material Generation with Diffusion Models (CVPR2024)
Shilin-LU/TF-ICON
[ICCV 2023] "TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Composition" (Official...