JianfeiJ/MonoMVSNet

[ICCV 2025] MonoMVSNet: Monocular Priors Guided Multi-View Stereo Network

32
/ 100
Emerging

This tool helps 3D reconstruction specialists and computer vision engineers create detailed 3D models from multiple camera images. It takes a sequence of calibrated images of an object or scene and generates a dense 3D point cloud, even for challenging areas like shiny or smooth surfaces. This is ideal for professionals needing highly accurate 3D representations.

Use this if you need to generate precise 3D point clouds from multiple images, especially when dealing with objects or scenes that have difficult-to-reconstruct areas like textureless or reflective surfaces.

Not ideal if you only have a single image and need to estimate relative depth, as this tool is specifically designed for multi-view stereo reconstruction.

3D-reconstruction computer-vision photogrammetry dense-reconstruction scene-modeling
No Package No Dependents
Maintenance 6 / 25
Adoption 7 / 25
Maturity 15 / 25
Community 4 / 25

How are scores calculated?

Stars

31

Forks

1

Language

Python

License

MIT

Last pushed

Dec 16, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/JianfeiJ/MonoMVSNet"

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