TensoRF
OpenSfM
Our great sponsors
| TensoRF | OpenSfM | |
|---|---|---|
| 1 | 5 | |
| 1,132 | 3,219 | |
| - | 1.6% | |
| 2.9 | 7.6 | |
| 7 months ago | 11 days ago | |
| Python | Python | |
| MIT License | BSD 2-clause "Simplified" License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
TensoRF
-
Nvidia Research Turns 2D Photos into 3D Scenes in the Blink of an AI
This is great, and the paper+codebase they're referring to (but not linking, here [1]) is neat too.
The research is moving fast though, so if you want something almost as fast without specialized CUDA kernels (just plain pytorch) you're in luck: https://github.com/apchenstu/TensoRF
As a bonus you also get a more compact representation of the scene.
[1] https://github.com/NVlabs/instant-ngp
OpenSfM
-
Import many photogrammetry software's scenes into Blender
OpenSfM (JSON)
- Packages for Bundle Adjustment
- WebODM/OpenSfM 2D pixel to 3D point
-
Photogrammetry in Python
You're kidding right? There are lots of coders on github with open-sauce Python. The basic algorithm being used now is Structure-from-Motion (SfM). Photogrammetry started with projecting a red stripe light on the object, taking the photo, and manually stitching together common points. That uses edge detection, another algorithm. Everybody knows the Z-depth calculation algorithm, and that is starting to be done with AI. Sony is using some scrolling parallax algorithm with their photogrammetry software. It looks like Reality Capture can calculate the camera positions and angles first, then use photon tracing rather than ray tracing to determine the point cloud. OpenSfM https://github.com/mapillary/OpenSfM
-
What are some good libraries for structure-from-motion?
OpenSfM is written in Python and JS and pretty easy to understand.
What are some alternatives?
instant-ngp - Instant neural graphics primitives: lightning fast NeRF and more
colmap - COLMAP - Structure-from-Motion and Multi-View Stereo
nerf - Code release for NeRF (Neural Radiance Fields)
OpenMVG (open Multiple View Geometry) - open Multiple View Geometry library. Basis for 3D computer vision and Structure from Motion.
fill - Generative fill in 3D.
ZeroNet - ZeroNet - Decentralized websites using Bitcoin crypto and BitTorrent network
awesome-NeRF - A curated list of awesome neural radiance fields papers
gtsfm - End-to-end SFM framework based on GTSAM
Meshroom - 3D Reconstruction Software
TheiaSfM - An open source library for multiview geometry and structure from motion
Fingerprint-Feature-Extraction - Extract minutiae features from fingerprint images
face-movie - Automatically create a time-lapse morph sequence of a face using OpenCV and Dlib.