Revisit Self-supervised Depth Estimation with Local Structure-from-Motion
It addresses the disconnection between self-supervised depth estimation and SfM for computer vision researchers, offering an incremental improvement by integrating local SfM into self-supervision.
This work tackles the problem of self-supervised depth estimation by proposing a local Structure-from-Motion (SfM) scheme that jointly optimizes camera poses and depth adjustments using a novel bundle-RANSAC-adjustment algorithm, achieving benefits for state-of-the-art supervised depth and correspondence models within 5 frames.
Both self-supervised depth estimation and Structure-from-Motion (SfM) recover scene depth from RGB videos. Despite sharing a similar objective, the two approaches are disconnected. Prior works of self-supervision backpropagate losses defined within immediate neighboring frames. Instead of learning-through-loss, this work proposes an alternative scheme by performing local SfM. First, with calibrated RGB or RGB-D images, we employ a depth and correspondence estimator to infer depthmaps and pair-wise correspondence maps. Then, a novel bundle-RANSAC-adjustment algorithm jointly optimizes camera poses and one depth adjustment for each depthmap. Finally, we fix camera poses and employ a NeRF, however, without a neural network, for dense triangulation and geometric verification. Poses, depth adjustments, and triangulated sparse depths are our outputs. For the first time, we show self-supervision within $5$ frames already benefits SoTA supervised depth and correspondence models. The project page is held in the link (https://shngjz.github.io/SSfM.github.io/).