Multi-Model Learning for Real-Time Automotive Semantic Foggy Scene Understanding via Domain Adaptation

TL;DR

This paper introduces a real-time automotive scene understanding system that combines domain adaptation to clear foggy images with multi-modal semantic segmentation, improving visibility and accuracy under adverse weather conditions.

Contribution

The paper presents a novel end-to-end multi-model approach that transforms foggy images to clear conditions and performs efficient semantic segmentation with low computational cost.

Findings

Achieves real-time performance suitable for automotive use

Provides comparable accuracy to state-of-the-art methods

Effectively exploits RGB, depth, and luminance data for robust segmentation

Abstract

Robust semantic scene segmentation for automotive applications is a challenging problem in two key aspects: (1) labelling every individual scene pixel and (2) performing this task under unstable weather and illumination changes (e.g., foggy weather), which results in poor outdoor scene visibility. Such visibility limitations lead to non-optimal performance of generalised deep convolutional neural network-based semantic scene segmentation. In this paper, we propose an efficient end-to-end automotive semantic scene understanding approach that is robust to foggy weather conditions. As an end-to-end pipeline, our proposed approach provides: (1) the transformation of imagery from foggy to clear weather conditions using a domain transfer approach (correcting for poor visibility) and (2) semantically segmenting the scene using a competitive encoder-decoder architecture with low computational complexity (enabling real-time performance). Our approach incorporates RGB colour, depth and luminance images via distinct encoders with dense connectivity and features fusion to effectively exploit information from different inputs, which contributes to an optimal feature representation within the overall model. Using this architectural formulation with dense skip connections, our model achieves comparable performance to contemporary approaches at a fraction of the overall model complexity.