Abstract:In this paper, we study the problem of jointly estimating the optical flow and scene flow from synchronized 2D and 3D data. Previous methods either employ a complex pipeline that splits the joint task into independent stages, or fuse 2D and 3D information in an ``early-fusion'' or ``late-fusion'' manner. Such one-size-fits-all approaches suffer from a dilemma of failing to fully utilize the characteristic of each modality or to maximize the inter-modality complementarity. To address the problem, we propose a novel end-to-end framework, which consists of 2D and 3D branches with multiple bidirectional fusion connections between them in specific layers. Different from previous work, we apply a point-based 3D branch to extract the LiDAR features, as it preserves the geometric structure of point clouds. To fuse dense image features and sparse point features, we propose a learnable operator named bidirectional camera-LiDAR fusion module (Bi-CLFM). We instantiate two types of the bidirectional fusion pipeline, one based on the pyramidal coarse-to-fine architecture (dubbed CamLiPWC), and the other one based on the recurrent all-pairs field transforms (dubbed CamLiRAFT). On FlyingThings3D, both CamLiPWC and CamLiRAFT surpass all existing methods and achieve up to a 47.9\% reduction in 3D end-point-error from the best published result. Our best-performing model, CamLiRAFT, achieves an error of 4.26\% on the KITTI Scene Flow benchmark, ranking 1st among all submissions with much fewer parameters. Besides, our methods have strong generalization performance and the ability to handle non-rigid motion. Code is available at <a class="link-external link-https" href="https://github.com/MCG-NJU/CamLiFlow" rel="external noopener nofollow">this https URL</a>.

Camera Motion Estimation from RGB-D-Inertial Scene Flow

Unsupervised Learning of Scene Flow Estimation Fusing with Local Rigidity.

Recurrent Volume-based 3D Feature Fusion for Real-time Multi-view Object Pose Estimation

Motion Recovery Using the Image Interpolation Algorithm and an RGB-D Camera

Inertial-aided State and Slope Estimation Using a Monocular Camera

Salient-motion-heuristic Scheme for Fast 3D Optical Flow Estimation Using RGB-D Data

ScaleFlow++: Robust and Accurate Estimation of 3D Motion from Video

Inertial-Aided Metric States and Surface Normal Estimation Using a Monocular Camera.

RPEFlow: Multimodal Fusion of RGB-PointCloud-Event for Joint Optical Flow and Scene Flow Estimation

From Scene Flow to Visual Odometry Through Local and Global Regularisation in Markov Random Fields

A Homography-Based Visual Inertial Fusion Method for Robust Sensing of a Micro Aerial Vehicle

Motion Detection Based on RGB-D Data and Scene Flow Clustering

CamLiFlow: Bidirectional Camera-LiDAR Fusion for Joint Optical Flow and Scene Flow Estimation

Learning Optical Flow and Scene Flow with Bidirectional Camera-LiDAR Fusion

FlowFusion: Dynamic Dense RGB-D SLAM Based on Optical Flow

EMR-MSF: Self-Supervised Recurrent Monocular Scene Flow Exploiting Ego-Motion Rigidity

Scene Motion Decomposition for Learnable Visual Odometry

Real-time Indoor Scene Reconstruction with RGBD and Inertial Input.

Fast visual odometry and mapping from RGB-D data

Dense RGB-D-Inertial SLAM with Map Deformations