Salient Object Detection in Complex Weather Conditions via Noise Indicators

TL;DR

This paper introduces a weather-aware salient object detection framework that uses noise indicators and a specialized encoder to improve segmentation accuracy in adverse weather conditions, validated on the WXSOD dataset.

Contribution

The paper presents a novel SOD framework with a noise indicator and a dedicated encoder that effectively handles diverse weather-induced noise, enhancing robustness over existing methods.

Findings

Improved segmentation accuracy in complex weather conditions.

Effective integration of weather noise indicators into the SOD framework.

Compatibility of the proposed encoder with mainstream decoders.

Abstract

Salient object detection (SOD), a foundational task in computer vision, has advanced from single-modal to multi-modal paradigms to enhance generalization. However, most existing SOD methods assume low-noise visual conditions, overlooking the degradation of segmentation accuracy caused by weather-induced noise in real-world scenarios. In this paper, we propose a SOD framework tailored for diverse weather conditions, encompassing a specific encoder and a replaceable decoder. To enable handling of varying weather noises, we introduce a one-hot vector as a noise indicator to represent different weather types and design a Noise Indicator Fusion Module (NIFM). The NIFM takes both semantic features and the noise indicator as dual inputs and is inserted between consecutive stages of the encoder to embed weather-aware priors via adaptive feature modulation. Critically, the proposed specific encoder retains compatibility with mainstream SOD decoders. Extensive experiments are conducted on the WXSOD dataset under varying training data scales (100%, 50%, 30% of the full training set), three encoder and seven decoder configurations. Results show that the proposed SOD framework (particularly the NIFM-enhanced specific encoder) improves segmentation accuracy under complex weather conditions compared to a vanilla encoder.