SfMLearner
An unsupervised learning framework for depth and ego-motion estimation from monocular videos using TensorFlow.
What is SfMLearner?
SfMLearner is an unsupervised deep learning framework that simultaneously estimates depth maps from single images and camera ego-motion from video sequences. It solves the structure-from-motion problem using self-supervision from view synthesis, eliminating the need for labeled 3D data. The system was originally presented in a CVPR 2017 paper and achieves state-of-the-art results on KITTI benchmarks.
Computer vision researchers and engineers working on 3D scene understanding, particularly those focused on autonomous driving, robotics, or AR/VR applications requiring geometry estimation from monocular cameras.
Developers choose SfMLearner because it provides a complete, well-documented implementation of a seminal unsupervised learning approach for 3D vision, with pre-trained models and evaluation code for standard datasets. Its self-supervised paradigm reduces data annotation costs while maintaining competitive accuracy.
Overview
An unsupervised learning framework for depth and ego-motion estimation from monocular videos
Use Cases
Best For
- Research on unsupervised depth estimation from monocular video
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
