End-to-end Learning of Cost-Volume Aggregation for Real-time Dense Stereo

TL;DR

This paper introduces an end-to-end deep learning approach for dense stereo matching that combines classical matching scores with learned cost volume aggregation, achieving high accuracy and real-time performance.

Contribution

It presents a novel end-to-end trainable system that integrates classical matching scores with a deep network for cost volume aggregation, implemented as a differentiable domain transform.

Findings

Achieves 6.34% error rate on KITTI 2015 benchmark.

Runs at 29 frames per second on a modern GPU.

Does not rely on deep learning of pixel appearance descriptors.

Abstract

We present a new deep learning-based approach for dense stereo matching. Compared to previous works, our approach does not use deep learning of pixel appearance descriptors, employing very fast classical matching scores instead. At the same time, our approach uses a deep convolutional network to predict the local parameters of cost volume aggregation process, which in this paper we implement using differentiable domain transform. By treating such transform as a recurrent neural network, we are able to train our whole system that includes cost volume computation, cost-volume aggregation (smoothing), and winner-takes-all disparity selection end-to-end. The resulting method is highly efficient at test time, while achieving good matching accuracy. On the KITTI 2015 benchmark, it achieves a result of 6.34\% error rate while running at 29 frames per second rate on a modern GPU.