M2Restore: Mixture-of-Experts-based Mamba-CNN Fusion Framework for All-in-One Image Restoration

TL;DR

M2Restore is a novel image restoration framework that combines mixture-of-experts, CLIP-guided gating, and dual-stream CNN-Mamba architecture to enhance generalization, detail preservation, and efficiency across diverse degradation scenarios.

Contribution

It introduces a CLIP-guided MoE gating mechanism, a dual-stream CNN-Mamba architecture, and an edge-aware dynamic gating for improved all-in-one image restoration.

Findings

Outperforms existing methods on multiple benchmarks

Enhances global coherence and local detail restoration

Achieves superior visual quality and quantitative metrics

Abstract

Natural images are often degraded by complex, composite degradations such as rain, snow, and haze, which adversely impact downstream vision applications. While existing image restoration efforts have achieved notable success, they are still hindered by two critical challenges: limited generalization across dynamically varying degradation scenarios and a suboptimal balance between preserving local details and modeling global dependencies. To overcome these challenges, we propose M2Restore, a novel Mixture-of-Experts (MoE)-based Mamba-CNN fusion framework for efficient and robust all-in-one image restoration. M2Restore introduces three key contributions: First, to boost the model's generalization across diverse degradation conditions, we exploit a CLIP-guided MoE gating mechanism that fuses task-conditioned prompts with CLIP-derived semantic priors. This mechanism is further refined via cross-modal feature calibration, which enables precise expert selection for various degradation types. Second, to jointly capture global contextual dependencies and fine-grained local details, we design a dual-stream architecture that integrates the localized representational strength of CNNs with the long-range modeling efficiency of Mamba. This integration enables collaborative optimization of global semantic relationships and local structural fidelity, preserving global coherence while enhancing detail restoration. Third, we introduce an edge-aware dynamic gating mechanism that adaptively balances global modeling and local enhancement by reallocating computational attention to degradation-sensitive regions. This targeted focus leads to more efficient and precise restoration. Extensive experiments across multiple image restoration benchmarks validate the superiority of M2Restore in both visual quality and quantitative performance.