Depth-Synergized Mamba Meets Memory Experts for All-Day Image Reflection Separation

2026-01-01

arXiv.org

AI Summary

This work addresses the challenge of separating transmission and reflection layers in a single image when their contrast is similar—particularly problematic in nighttime scenes. To this end, the authors propose the Depth-Memory Decoupling Network (DMDNet), which uniquely integrates depth information with the Mamba architecture. The method introduces a depth-aware scanning mechanism and a Depth-Synergized State Space Model (DS-SSM), along with a Memory Expert Compensation Module (MECM) to leverage historical knowledge across images. Additionally, the study presents NightIRS, the first benchmark dataset dedicated to nighttime reflection separation. Experimental results demonstrate that DMDNet significantly outperforms existing approaches in both daytime and nighttime scenarios, achieving enhanced accuracy and robustness in layer separation.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Image reflection separation aims to disentangle the transmission layer and the reflection layer from a blended image. Existing methods rely on limited information from a single image, tending to confuse the two layers when their contrasts are similar, a challenge more severe at night. To address this issue, we propose the Depth-Memory Decoupling Network (DMDNet). It employs the Depth-Aware Scanning (DAScan) to guide Mamba toward salient structures, promoting information flow along semantic coherence to construct stable states. Working in synergy with DAScan, the Depth-Synergized State-Space Model (DS-SSM) modulates the sensitivity of state activations by depth, suppressing the spread of ambiguous features that interfere with layer disentanglement. Furthermore, we introduce the Memory Expert Compensation Module (MECM), leveraging cross-image historical knowledge to guide experts in providing layer-specific compensation. To address the lack of datasets for nighttime reflection separation, we construct the Nighttime Image Reflection Separation (NightIRS) dataset. Extensive experiments demonstrate that DMDNet outperforms state-of-the-art methods in both daytime and nighttime.

Problem

• image reflection separation
• nighttime
• layer disentanglement
• contrast similarity
• transmission and reflection layers

Innovation

• Depth-Synergized Mamba
• Memory Experts
• Reflection Separation
• State-Space Model
• Nighttime Image Dataset