baegwangbin/DSINE
[CVPR 2024 Oral] Rethinking Inductive Biases for Surface Normal Estimation
This tool helps improve the accuracy and crispness of surface normal estimations from images, a crucial step for 3D reconstruction and understanding object geometry. It takes various images (from samples, webcams, or video feeds) and camera intrinsic data, then outputs refined surface normal maps. This is ideal for researchers and practitioners working on computer vision tasks that require highly precise surface geometry from visual data.
893 stars. No commits in the last 6 months.
Use this if you need to generate exceptionally crisp and accurate surface normal maps from images, even challenging 'in-the-wild' ones, and want a method that generalizes well despite smaller training datasets.
Not ideal if your primary goal is general-purpose dense prediction, as this tool is specifically optimized for surface normal estimation with tailored inductive biases.
Stars
893
Forks
43
Language
Jupyter Notebook
License
—
Category
Last pushed
Jul 10, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/baegwangbin/DSINE"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
3DOM-FBK/deep-image-matching
Multiview matching with deep-learning and hand-crafted local features for COLMAP and other SfM...
suhangpro/mvcnn
Multi-view CNN (MVCNN) for shape recognition
zouchuhang/LayoutNet
Torch implementation of our CVPR 18 paper: "LayoutNet: Reconstructing the 3D Room Layout from a...
andyzeng/tsdf-fusion-python
Python code to fuse multiple RGB-D images into a TSDF voxel volume.
andyzeng/tsdf-fusion
Fuse multiple depth frames into a TSDF voxel volume.