cvlab-columbia/pix2gestalt

Code for the paper "pix2gestalt: Amodal Segmentation by Synthesizing Wholes" (CVPR 2024)

38
/ 100
Emerging

When working with images where objects are partially hidden, this tool helps you understand and process those objects as if they were fully visible. It takes an image with occluded objects and, using your input to outline the visible parts, it outputs a complete, unoccluded image of each object, along with precise boundaries for both the visible and imagined hidden parts. This is for computer vision researchers, robotics engineers, or anyone developing systems that need to 'see' and interact with objects even when they're partially blocked.

200 stars. No commits in the last 6 months.

Use this if you need to perform object recognition, scene understanding, or 3D reconstruction on images containing objects that are partially obscured.

Not ideal if you are looking for a tool that works in real-time on consumer-grade hardware or if your primary interest is simple visible-only object segmentation.

computer-vision-research robotics scene-understanding amodal-perception object-recognition
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 10 / 25

How are scores calculated?

Stars

200

Forks

11

Language

Python

License

Last pushed

Jun 26, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/cvlab-columbia/pix2gestalt"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.