SynDistNet: Self-Supervised Monocular Fisheye Camera Distance Estimation Synergized with Semantic Segmentation for Autonomous Driving

TL;DR

This paper presents SynDistNet, a self-supervised multi-task learning approach that enhances monocular distance estimation for fisheye and pinhole cameras in autonomous driving, addressing scale ambiguity and dynamic object artifacts.

Contribution

Introduces a novel self-attention based network with semantic guidance and a robust loss, improving monocular distance estimation for fisheye cameras without external scale.

Findings

25% reduction in RMSE over previous fisheye methods

State-of-the-art results on self-supervised monocular distance estimation

Effective handling of dynamic objects with semantic masking

Abstract

State-of-the-art self-supervised learning approaches for monocular depth estimation usually suffer from scale ambiguity. They do not generalize well when applied on distance estimation for complex projection models such as in fisheye and omnidirectional cameras. This paper introduces a novel multi-task learning strategy to improve self-supervised monocular distance estimation on fisheye and pinhole camera images. Our contribution to this work is threefold: Firstly, we introduce a novel distance estimation network architecture using a self-attention based encoder coupled with robust semantic feature guidance to the decoder that can be trained in a one-stage fashion. Secondly, we integrate a generalized robust loss function, which improves performance significantly while removing the need for hyperparameter tuning with the reprojection loss. Finally, we reduce the artifacts caused by dynamic objects violating static world assumptions using a semantic masking strategy. We significantly improve upon the RMSE of previous work on fisheye by 25% reduction in RMSE. As there is little work on fisheye cameras, we evaluated the proposed method on KITTI using a pinhole model. We achieved state-of-the-art performance among self-supervised methods without requiring an external scale estimation.