EdgeConv with Attention Module for Monocular Depth Estimation

TL;DR

This paper introduces novel EdgeConv-based modules with attention mechanisms to improve monocular depth estimation, achieving state-of-the-art results on NYU Depth V2 and KITTI datasets, especially in challenging conditions.

Contribution

The paper presents a new Patch-Wise EdgeConv Module and EdgeConv Attention Module that enhance structural learning for monocular depth estimation.

Findings

Achieves state-of-the-art performance on NYU Depth V2 dataset.

Demonstrates robustness in challenging scenes with various experiments.

Improves depth prediction accuracy using edge convolution and attention mechanisms.

Abstract

Monocular depth estimation is an especially important task in robotics and autonomous driving, where 3D structural information is essential. However, extreme lighting conditions and complex surface objects make it difficult to predict depth in a single image. Therefore, to generate accurate depth maps, it is important for the model to learn structural information about the scene. We propose a novel Patch-Wise EdgeConv Module (PEM) and EdgeConv Attention Module (EAM) to solve the difficulty of monocular depth estimation. The proposed modules extract structural information by learning the relationship between image patches close to each other in space using edge convolution. Our method is evaluated on two popular datasets, the NYU Depth V2 and the KITTI Eigen split, achieving state-of-the-art performance. We prove that the proposed model predicts depth robustly in challenging scenes through various comparative experiments.