Exploring Self-Attention for Visual Odometry

TL;DR

This paper investigates the use of self-attention mechanisms in visual odometry networks to improve feature extraction and odometry accuracy, especially in scenes with dynamic objects and texture-less regions.

Contribution

It introduces self-attention into visual odometry models and demonstrates its effectiveness through qualitative, quantitative, and saliency-based analyses.

Findings

Self-attention improves feature quality for odometry.

Enhanced odometry accuracy over state-of-the-art methods.

Saliency studies show better focus on relevant regions.

Abstract

Visual odometry networks commonly use pretrained optical flow networks in order to derive the ego-motion between consecutive frames. The features extracted by these networks represent the motion of all the pixels between frames. However, due to the existence of dynamic objects and texture-less surfaces in the scene, the motion information for every image region might not be reliable for inferring odometry due to the ineffectiveness of dynamic objects in derivation of the incremental changes in position. Recent works in this area lack attention mechanisms in their structures to facilitate dynamic reweighing of the feature maps for extracting more refined egomotion information. In this paper, we explore the effectiveness of self-attention in visual odometry. We report qualitative and quantitative results against the SOTA methods. Furthermore, saliency-based studies alongside specially designed experiments are utilized to investigate the effect of self-attention on VO. Our experiments show that using self-attention allows for the extraction of better features while achieving a better odometry performance compared to networks that lack such structures.