URWKV: Unified RWKV Model with Multi-state Perspective for Low-light Image Restoration

2025-05-29

AI Summary

Existing low-light image enhancement (LLIE) and joint LLIE-deblurring methods suffer from limited modeling capacity and poor generalization under dynamic coupled degradations—e.g., concurrent low illumination and motion blur. To address this, we propose a lightweight, unified restoration framework leveraging RWKV-based sequential modeling. Our key contributions are: (1) Lightness-Adaptive Normalization (LAN), which emulates human pupillary response to improve illumination robustness; (2) a multi-stage state-aggregation mechanism using exponential moving averages, enabling both cross-stage feature alignment and intra-stage temporal modeling; and (3) a State-Sensitive Selective Fusion (SSF) module that replaces static skip connections to enhance structural adaptability. Extensive experiments demonstrate substantial improvements over state-of-the-art methods across multiple benchmarks. Our approach reduces parameter count and FLOPs by over 30%, while achieving superior detail recovery and enhanced adaptability to complex, dynamically degraded scenes.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Large Vision Models
• Search and Optimization: Learning to Search

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
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Existing low-light image enhancement (LLIE) and joint LLIE and deblurring (LLIE-deblur) models have made strides in addressing predefined degradations, yet they are often constrained by dynamically coupled degradations. To address these challenges, we introduce a Unified Receptance Weighted Key Value (URWKV) model with multi-state perspective, enabling flexible and effective degradation restoration for low-light images. Specifically, we customize the core URWKV block to perceive and analyze complex degradations by leveraging multiple intra- and inter-stage states. First, inspired by the pupil mechanism in the human visual system, we propose Luminance-adaptive Normalization (LAN) that adjusts normalization parameters based on rich inter-stage states, allowing for adaptive, scene-aware luminance modulation. Second, we aggregate multiple intra-stage states through exponential moving average approach, effectively capturing subtle variations while mitigating information loss inherent in the single-state mechanism. To reduce the degradation effects commonly associated with conventional skip connections, we propose the State-aware Selective Fusion (SSF) module, which dynamically aligns and integrates multi-state features across encoder stages, selectively fusing contextual information. In comparison to state-of-the-art models, our URWKV model achieves superior performance on various benchmarks, while requiring significantly fewer parameters and computational resources.

Problem

• Address dynamically coupled degradations in low-light images
• Enhance and deblur low-light images adaptively
• Reduce information loss in single-state mechanisms

Innovation

• Luminance-adaptive Normalization for scene-aware modulation
• Exponential moving average captures subtle variations
• State-aware Selective Fusion integrates multi-state features