Single image depth estimation by dilated deep residual convolutional neural network and soft-weight-sum inference

TL;DR

This paper introduces a novel residual CNN architecture utilizing dilated convolution, skip connections, and soft-weight-sum inference for single image depth estimation, achieving superior results with fewer data and parameters.

Contribution

The paper presents a new residual CNN design that improves depth estimation accuracy while reducing training data and model complexity.

Findings

Outperforms state-of-the-art methods on NYU Depth V2 dataset

Uses fewer training examples and model parameters

Achieves better depth estimation accuracy

Abstract

This paper proposes a new residual convolutional neural network (CNN) architecture for single image depth estimation. Compared with existing deep CNN based methods, our method achieves much better results with fewer training examples and model parameters. The advantages of our method come from the usage of dilated convolution, skip connection architecture and soft-weight-sum inference. Experimental evaluation on the NYU Depth V2 dataset shows that our method outperforms other state-of-the-art methods by a margin.