Practical Deep Stereo (PDS): Toward applications-friendly deep stereo matching

Abstract

End-to-end deep-learning networks recently demonstrated extremely good perfor- mance for stereo matching. However, existing networks are difficult to use for practical applications since (1) they are memory-hungry and unable to process even modest-size images, (2) they have to be trained for a given disparity range. The Practical Deep Stereo (PDS) network that we propose addresses both issues: First, its architecture relies on novel bottleneck modules that drastically reduce the memory footprint in inference, and additional design choices allow to handle greater image size during training. This results in a model that leverages large image context to resolve matching ambiguities. Second, a novel sub-pixel cross- entropy loss combined with a MAP estimator make this network less sensitive to ambiguous matches, and applicable to any disparity range without re-training. We compare PDS to state-of-the-art methods published over the recent months, and demonstrate its superior performance on FlyingThings3D and KITTI sets.

Cited by in corpus (18)

• A Survey on Deep Learning Techniques for Stereo-based Depth Estimation
• AMNet: Deep Atrous Multiscale Stereo Disparity Estimation Networks
• GA-Net: Guided Aggregation Net for End-to-end Stereo Matching
• Learning to Think Outside the Box: Wide-Baseline Light Field Depth Estimation with EPI-Shift
• A Survey on Deep Learning Architectures for Image-based Depth Reconstruction
• Hierarchical Discrete Distribution Decomposition for Match Density Estimation
• Adaptive Unimodal Cost Volume Filtering for Deep Stereo Matching
• EdgeStereo: An Effective Multi-Task Learning Network for Stereo Matching and Edge Detection
• Group-wise Correlation Stereo Network
• Improving Deep Stereo Network Generalization with Geometric Priors
• Learning Inverse Depth Regression for Multi-View Stereo with Correlation Cost Volume
• Hierarchical Deep Stereo Matching on High-resolution Images
• Content-Aware Inter-Scale Cost Aggregation for Stereo Matching
• Attention Aware Cost Volume Pyramid Based Multi-view Stereo Network for 3D Reconstruction
• Increased-Range Unsupervised Monocular Depth Estimation
• ADCPNet: Adaptive Disparity Candidates Prediction Network for Efficient Real-Time Stereo Matching
• LSM: Learning Subspace Minimization for Low-level Vision
• On the Synergies between Machine Learning and Binocular Stereo for Depth Estimation from Images: a Survey