Every Pixel Counts ++: Joint Learning of Geometry and Motion with 3D Holistic Understanding

Abstract

Learning to estimate 3D geometry in a single frame and optical flow from consecutive frames by watching unlabeled videos via deep convolutional network has made significant progress recently. Current state-of-the-art (SoTA) methods treat the two tasks independently. One typical assumption of the existing depth estimation methods is that the scenes contain no independent moving objects. while object moving could be easily modeled using optical flow. In this paper, we propose to address the two tasks as a whole, i.e. to jointly understand per-pixel 3D geometry and motion. This eliminates the need of static scene assumption and enforces the inherent geometrical consistency during the learning process, yielding significantly improved results for both tasks. We call our method as "Every Pixel Counts++" or "EPC++". Specifically, during training, given two consecutive frames from a video, we adopt three parallel networks to predict the camera motion (MotionNet), dense depth map (DepthNet), and per-pixel optical flow between two frames (OptFlowNet) respectively. The three types of information are fed into a holistic 3D motion parser (HMP), and per-pixel 3D motion of both rigid background and moving objects are disentangled and recovered. Comprehensive experiments were conducted on datasets with different scenes, including driving scenario (KITTI 2012 and KITTI 2015 datasets), mixed outdoor/indoor scenes (Make3D) and synthetic animation (MPI Sintel dataset). Performance on the five tasks of depth estimation, optical flow estimation, odometry, moving object segmentation and scene flow estimation shows that our approach outperforms other SoTA methods. Code will be available at: https://github.com/chenxuluo/EPC.

arXiv commentChenxu Luo, Zhenheng Yang, and Peng Wang contributed equally, TPAMI submission

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
• DepthTransfer: Depth Extraction from Video Using Non-parametric Sampling
• SfM-Net: Learning of Structure and Motion from Video
• Semi-Supervised Deep Learning for Monocular Depth Map Prediction

Cited by in corpus (8)

• On Deep Learning Techniques to Boost Monocular Depth Estimation for Autonomous Navigation
• Monocular Depth Estimation with Self-supervised Instance Adaptation
• Feature-metric Loss for Self-supervised Learning of Depth and Egomotion
• Self-Supervised Pillar Motion Learning for Autonomous Driving
• Calibrating Self-supervised Monocular Depth Estimation
• Single Image Depth Prediction with Wavelet Decomposition
• Domain Adaptive Monocular Depth Estimation With Semantic Information
• Semantics-Driven Unsupervised Learning for Monocular Depth and Ego-Motion Estimation