Semantic Segmentation under Adverse Conditions: A Weather and Nighttime-aware Synthetic Data-based Approach

TL;DR

This paper introduces a weather and nighttime-aware extension to DeepLabV3+ that leverages synthetic data for improved semantic segmentation under adverse conditions, achieving significant accuracy gains without sacrificing performance in normal weather.

Contribution

A novel architecture with multi-task learning for weather and time-of-day awareness, enhancing domain adaptation using synthetic data for adverse condition segmentation.

Findings

Achieves 14% mIoU improvement on ACDC dataset.

Maintains 75% mIoU on Cityscapes dataset.

Demonstrates effective domain adaptation with synthetic data.

Abstract

Recent semantic segmentation models perform well under standard weather conditions and sufficient illumination but struggle with adverse weather conditions and nighttime. Collecting and annotating training data under these conditions is expensive, time-consuming, error-prone, and not always practical. Usually, synthetic data is used as a feasible data source to increase the amount of training data. However, just directly using synthetic data may actually harm the model's performance under normal weather conditions while getting only small gains in adverse situations. Therefore, we present a novel architecture specifically designed for using synthetic training data for domain adaptation. We propose a simple yet powerful addition to DeepLabV3+ by using weather and time-of-the-day supervisors trained with multi-task learning, making it both weather and nighttime aware, which improves its mIoU accuracy by $14$ percentage points on the ACDC dataset while maintaining a score of $75\%$ mIoU on the Cityscapes dataset. Our code is available at https://github.com/lsmcolab/Semantic-Segmentation-under-Adverse-Conditions.