BEVScope: Enhancing Self-Supervised Depth Estimation Leveraging Bird's-Eye-View in Dynamic Scenarios

TL;DR

BEVScope introduces a novel self-supervised depth estimation method that leverages Bird's-Eye-View features and an adaptive loss to improve performance in dynamic scenarios, especially with moving objects.

Contribution

The paper proposes BEVScope, a new approach that effectively utilizes BEV features and an adaptive loss function for better depth estimation in dynamic environments.

Findings

Demonstrates competitive performance on Nuscenes dataset.

Effectively handles dynamic objects with the proposed adaptive loss.

Leverages multi-camera views through BEV features.

Abstract

Depth estimation is a cornerstone of perception in autonomous driving and robotic systems. The considerable cost and relatively sparse data acquisition of LiDAR systems have led to the exploration of cost-effective alternatives, notably, self-supervised depth estimation. Nevertheless, current self-supervised depth estimation methods grapple with several limitations: (1) the failure to adequately leverage informative multi-camera views. (2) the limited capacity to handle dynamic objects effectively. To address these challenges, we present BEVScope, an innovative approach to self-supervised depth estimation that harnesses Bird's-Eye-View (BEV) features. Concurrently, we propose an adaptive loss function, specifically designed to mitigate the complexities associated with moving objects. Empirical evaluations conducted on the Nuscenes dataset validate our approach, demonstrating competitive performance. Code will be released at https://github.com/myc634/BEVScope.