Abstract:Scene flow estimation in the dynamic scene remains a challenging task. Computing scene flow by a combination of 2D optical flow and depth has shown to be considerably faster with acceptable performance. In this work, we present a unified framework for joint unsupervised learning of stereo depth and optical flow with explicit local rigidity to estimate scene flow. We estimate camera motion directly by a Perspective-n-Point method from the optical flow and depth predictions, with RANSAC outlier rejection scheme. In order to disambiguate the object motion and the camera motion in the scene, we distinguish the rigid region by the re-project error and the photometric similarity. By joint learning with the local rigidity, both depth and optical networks can be refined. This framework boosts all four tasks: depth, optical flow, camera motion estimation, and object motion segmentation. Through the evaluation on the KITTI benchmark, we show that the proposed framework achieves state-of-the-art results amongst unsupervised methods. Our models and code are available at https://github.com/lliuz/unrigidflow.

Linear Model for 3d Motion Estimation and Shape Reconstruction Based on the Straight-Line Optical Flow

A New Non-Linear Algorithm for 3D Reconstruction from Straight-line Optical Flow

Three-dimensional motion estimation by straight-line and endpoint optical flow

Reconstructing 3D Motion and Structure from Straight-Line Optical Flow

Stereo Camera Motion Estimation Based on Euclidean Reconstruction

Experiment Simulation of 3D Reconstruction Linear Algorithm Based on Optical Flow

Direct Method for Motion Estimation and Structure Reconstruction Based on Optical Flow

Unsupervised Learning of Scene Flow Estimation Fusing with Local Rigidity.

A New Method of 3D Motion and Structure Estimation Based on Spare Optical Flow

Calculation of 3D Motion and Structure from Optical Flow in Image Sequence

3d Motion Segmentation From Straight-Line Optical Flow

A Variational Method for Recovering 3D Structure of Purely Rotational Object from Image Sequence

A Closed Form Solution to Segment 3D Motion Using Straight-Line Optical Flow

A Linear Algorithm for Motion Estimation Using Straight Line Correspondences

A NONLINEAR FILTERING BASED OPTICAL FLOW COMPUTATION

Optical Flow Method for Motion Details Estimation

Optical Flow-based 3D Human Motion Estimation from Monocular Video

A Fast Simple Optical Flow Computation Approach Based on the 3-D Gradient

Linear Motion Estimation Using Straight-Line Correspondences And Its Error Analysis

An Optical Flow Model Based on Illumination Radiation Theory

Recovery Of Egomotion From Optical Flow With Large Motion Based On Subspace Method