Salient Object Detection in Complex Weather Conditions via Noise Indicators

2025-12-11

AI Summary

To address the significant degradation in salient object detection (SOD) performance under complex weather conditions—such as rain, fog, and snow—this paper proposes a noise-aware SOD framework. Methodologically, it introduces three key innovations: (i) a novel one-hot-encoded explicit weather-type indicator to encode weather priors; (ii) a plug-and-play Noise Indicator Fusion Module (NIFM) that adaptively fuses weather priors with semantic features; and (iii) a multi-scale feature modulation mechanism coupled with cross-weather generalization training. The framework is compatible with mainstream CNN-based decoders. Evaluated on the WXSOD benchmark, it consistently outperforms state-of-the-art baselines, achieving an average F-measure gain of 2.1–3.8 percentage points. Notably, even when trained on only 30% of the full dataset, the method retains robust performance improvements, demonstrating strong generalization capability and data efficiency.

Technology Category

• Computer Vision: Object Detection & Categorization
• Machine Learning: Neuro-Symbolic Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

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.

Problem

• Detects salient objects in complex weather conditions
• Addresses segmentation accuracy degradation from weather noise
• Enhances generalization with noise-aware feature modulation

Innovation

• Noise indicator vector represents weather types
• NIFM module fuses semantic features with noise indicator
• Specific encoder is compatible with various SOD decoders