Depth estimation
lastmod 2026-11-10created 2020-05-073 edits
Depth estimation is a computer vision task that estimates how far each pixel in a 2D image or video is from the camera. Approaches include stereo methods, which use the disparity between two cameras, and monocular methods, which predict depth from a single image using deep learning; for video, the challenge is keeping depth consistent across frames. It is used in 3D reconstruction, autonomous driving, AR, and more.
Deep learning based
- Consistent Video Depth Estimation
- https://www.reddit.com/r/MachineLearning/comments/gc2wo9/r_consistent_video_depth_estimation_siggraph_2020/
- https://arxiv.org/abs/2004.15021
- https://roxanneluo.github.io/Consistent-Video-Depth-Estimation/