CFNet: Cascade and Fused Cost Volume for Robust Stereo Matching

Abstract

Recently, the ever-increasing capacity of large-scale annotated datasets has led to profound progress in stereo matching. However, most of these successes are limited to a specific dataset and cannot generalize well to other datasets. The main difficulties lie in the large domain differences and unbalanced disparity distribution across a variety of datasets, which greatly limit the real-world applicability of current deep stereo matching models. In this paper, we propose CFNet, a Cascade and Fused cost volume based network to improve the robustness of the stereo matching network. First, we propose a fused cost volume representation to deal with the large domain difference. By fusing multiple low-resolution dense cost volumes to enlarge the receptive field, we can extract robust structural representations for initial disparity estimation. Second, we propose a cascade cost volume representation to alleviate the unbalanced disparity distribution. Specifically, we employ a variance-based uncertainty estimation to adaptively adjust the next stage disparity search space, in this way driving the network progressively prune out the space of unlikely correspondences. By iteratively narrowing down the disparity search space and improving the cost volume resolution, the disparity estimation is gradually refined in a coarse-to-fine manner. When trained on the same training images and evaluated on KITTI, ETH3D, and Middlebury datasets with the fixed model parameters and hyperparameters, our proposed method achieves the state-of-the-art overall performance and obtains the 1st place on the stereo task of Robust Vision Challenge 2020. The code will be available at https://github.com/gallenszl/CFNet.

arXiv commentaccepted by CVPR2021

References in corpus (12)

• Understanding the Effective Receptive Field in Deep Convolutional Neural Networks
• End-to-End Learning of Geometry and Context for Deep Stereo Regression
• DeepDriving: Learning Affordance for Direct Perception in Autonomous Driving
• Self-Supervised Learning for Stereo Matching with Self-Improving Ability
• Pyramid Stereo Matching Network
• Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching
• Learning Depth with Convolutional Spatial Propagation Network
• Learning for Disparity Estimation through Feature Constancy
• Deep Stereo using Adaptive Thin Volume Representation with Uncertainty Awareness
• Hierarchical Discrete Distribution Decomposition for Match Density Estimation
• Improving Deep Stereo Network Generalization with Geometric Priors
• Domain-invariant Stereo Matching Networks