A Novel Monocular Disparity Estimation Network with Domain Transformation and Ambiguity Learning

TL;DR

This paper introduces a new unsupervised monocular disparity estimation network that improves accuracy, reduces parameters, and estimates full disparity maps in a single pass, outperforming previous methods on the KITTI dataset.

Contribution

The paper presents a novel encoder-decoder architecture with domain transformation and ambiguity learning for unsupervised monocular disparity estimation, achieving superior performance with fewer parameters.

Findings

Outperforms Monodepth baseline in all metrics

Reduces model parameters significantly

Estimates full disparity map in a single forward pass

Abstract

Convolutional neural networks (CNN) have shown state-of-the-art results for low-level computer vision problems such as stereo and monocular disparity estimations, but still, have much room to further improve their performance in terms of accuracy, numbers of parameters, etc. Recent works have uncovered the advantages of using an unsupervised scheme to train CNN's to estimate monocular disparity, where only the relatively-easy-to-obtain stereo images are needed for training. We propose a novel encoder-decoder architecture that outperforms previous unsupervised monocular depth estimation networks by (i) taking into account ambiguities, (ii) efficient fusion between encoder and decoder features with rectangular convolutions and (iii) domain transformations between encoder and decoder. Our architecture outperforms the Monodepth baseline in all metrics, even with a considerable reduction of parameters. Furthermore, our architecture is capable of estimating a full disparity map in a single forward pass, whereas the baseline needs two passes. We perform extensive experiments to verify the effectiveness of our method on the KITTI dataset.