Deep Ordinal Regression Network for Monocular Depth Estimation

TL;DR

This paper introduces a novel ordinal regression approach with a multi-scale network for monocular depth estimation, achieving state-of-the-art accuracy and faster convergence by reducing low-resolution features and simplifying training.

Contribution

The paper proposes a spacing-increasing discretization strategy and a multi-scale network to improve depth estimation accuracy and training efficiency.

Findings

Achieves state-of-the-art results on four benchmarks.

Faster convergence compared to traditional regression methods.

Wins first place in the 2018 Robust Vision Challenge.

Abstract

Monocular depth estimation, which plays a crucial role in understanding 3D scene geometry, is an ill-posed problem. Recent methods have gained significant improvement by exploring image-level information and hierarchical features from deep convolutional neural networks (DCNNs). These methods model depth estimation as a regression problem and train the regression networks by minimizing mean squared error, which suffers from slow convergence and unsatisfactory local solutions. Besides, existing depth estimation networks employ repeated spatial pooling operations, resulting in undesirable low-resolution feature maps. To obtain high-resolution depth maps, skip-connections or multi-layer deconvolution networks are required, which complicates network training and consumes much more computations. To eliminate or at least largely reduce these problems, we introduce a spacing-increasing discretization (SID) strategy to discretize depth and recast depth network learning as an ordinal regression problem. By training the network using an ordinary regression loss, our method achieves much higher accuracy and \dd{faster convergence in synch}. Furthermore, we adopt a multi-scale network structure which avoids unnecessary spatial pooling and captures multi-scale information in parallel. The method described in this paper achieves state-of-the-art results on four challenging benchmarks, i.e., KITTI [17], ScanNet [9], Make3D [50], and NYU Depth v2 [42], and win the 1st prize in Robust Vision Challenge 2018. Code has been made available at: https://github.com/hufu6371/DORN.