arXiv · 2112.01502
Dimensions of Motion: Monocular Prediction through Flow Subspaces
Abstract
We introduce a way to learn to estimate a scene representation from a single image by predicting a low-dimensional subspace of optical flow for each training example, which encompasses the variety of possible camera and object movement. Supervision is provided by a novel loss which measures the distance between this predicted flow subspace and an observed optical flow. This provides a new approach to learning scene representation tasks, such as monocular depth prediction or instance segmentation, in an unsupervised fashion using in-the-wild input videos without requiring camera poses, intrinsics, or an explicit multi-view stereo step. We evaluate our method in multiple settings, including an indoor depth prediction task where it achieves comparable performance to recent methods trained with more supervision.
Explore related subjects
Keep this discovery
Richard Strong Bowen, Richard Tucker, Ramin Zabih, Noah Snavely. 2021-12-02. Dimensions of Motion: Monocular Prediction through Flow Subspaces. https://arxiv.org/abs/2112.01502
Cite the original work for its findings. Save a collection to share your selection of sources.