ayoussf/VMatcher
VMatcher: State-Space Semi-Dense Local Feature Matching
This project helps computer vision engineers and researchers accurately find corresponding points between two different images, even when views or lighting change significantly. It takes two input images and outputs precise 'matches' between features in them. This is crucial for tasks like 3D reconstruction, augmented reality, or visual localization, where understanding how two images relate spatially is key.
No commits in the last 6 months.
Use this if you need a robust and computationally efficient way to identify shared visual elements between different images for tasks like 3D scene understanding or visual tracking.
Not ideal if your primary goal is object detection or image classification, as this tool focuses specifically on finding precise correspondences between image features rather than identifying entire objects or scenes.
Stars
17
Forks
—
Language
Python
License
MIT
Category
Last pushed
Aug 16, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/ayoussf/VMatcher"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
cvg/LightGlue
LightGlue: Local Feature Matching at Light Speed (ICCV 2023)
shariqfarooq123/AdaBins
Official implementation of Adabins: Depth Estimation using adaptive bins
amusi/CVPR2026-Papers-with-Code
CVPR 2026 论文和开源项目合集
isl-org/ZoeDepth
Metric depth estimation from a single image
daviddmc/NeSVoR
NeSVoR is a package for GPU-accelerated slice-to-volume reconstruction.