DA-RAW: Domain Adaptive Object Detection for Real-World Adverse Weather Conditions

Despite the success of deep learning-based object detection methods in recent years, it is still challenging to make the object detector reliable in adverse weather conditions such as rain and snow. For the robust performance of object detectors, unsupervised domain adaptation has been utilized to adapt the detection network trained on clear weather images to adverse weather images.

While previous methods do not explicitly address weather corruption during adaptation, the domain gap between clear and adverse weather can be decomposed into two factors with distinct characteristics: a style gap and a weather gap.

In this paper, we present an unsupervised domain adaptation framework for object detection that can more effectively adapt to real-world environments with adverse weather conditions by addressing these two gaps separately. Our method resolves the style gap by concentrating on stylerelated information of high-level features using an attention module. Using self-supervised contrastive learning, our framework then reduces the weather gap and acquires instance features that are robust to weather corruption. Extensive experiments demonstrate that our method outperforms other methods for object detection in adverse weather conditions.

▲ We resolve the style and weather gaps separately to achieve optimal feature alignment. To bridge the style gap, our method aligns high-level style-related features using an attention module. Moreover, self-supervised contrastive learning is employed to resolve the weather gap.

▲ In both rainy and snowy conditions, our approach outperforms other methods when applied to real-world datasets. Existing UDA methods such as SADA and SWDA show a significant improvement in performance on synthetic datasets, but their performance on real-world datasets is only marginally improved or even decreases. This implies that existing methods that globally align distributions are ineffective when adapting to real-world datasets with a large style gap and severe weather corruption.