Loop-Mamba: A Loop Mamba with Degradation-Aware and Shared Memory for Old Photo Restoration
2026-08-03 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed Loop-Mamba, a new method to fix old photos that have problems like scratches, fading, and blur by treating restoration as a step-by-step improvement process. They use a special tool called SGDE to understand the different types of damage in the photo and guide the fixing process. Loop-Mamba keeps track of the photo's details over multiple steps using a memory component, which helps it rebuild the picture more accurately without requiring heavy computation. They also created a new way to measure how well the restoration works, called ODRS. Tests show their method outperforms earlier approaches on standard benchmarks.
old photo restorationstate-space frameworkSemantic-Guided Degradation Estimatorstructural memoryiterative computationlatent restoration statemulti-directional scanningOld Photo Damage Recovery Scoregradient dilutionSynOld benchmark
Authors
Runci Bai, Yucheng Xin, Pu Wang, Yongcong Wang, Chen Wu, Dianjie Lu, Guijuan Zhang, Pengwen Dai, Guangwei Gao, Siyuan Yao, Zhuoran Zheng
Abstract
Old photographs often suffer from multiple coupled degradations, including scratches, cracks, fading, blur, noise, and missing regions, severely degrading both visual quality and semantic content. We propose Loop-Mamba, a lightweight loop-based state-space framework that formulates old photo restoration as progressive state evolution, where a persis- tent restoration state is continuously propagated and refined through iterative computation. Specifically, we introduce a Semantic-Guided Degradation Estimator (SGDE) to explicitly model heterogeneous degradations by jointly predicting local degradation maps and global degradation scores, providing degradation-aware guidance for state evolution. We further develop a Shared Structural Memory Mamba (S$^2$M- Mamba), which maintains a persistent restoration state across iterations, enabling persistent state evolution through shared structural memory for robust long-range structural reconstruction. Benefiting from first-order state recursion, Loop-Mamba propagates latent restoration states through recurrent tran- sitions instead of repeatedly stacking deep feature transformations, thereby alleviating gradient dilution while avoiding the computational overhead inherent in iterative CNN- and Transformer-based restoration frameworks. A lightweight multi-directional scanning strategy further enhances direc- tional information aggregation and preserves structural continuity. To better evaluate restoration quality, we introduce the task-oriented Old Photo Damage Recovery Score (ODRS), which jointly measures degradation recovery and structural reconstruction fidelity. Experimental results on the public SynOld benchmark demonstrate that Loop-Mamba consistently outperforms previous state-of-the-art methods across both conventional restoration metrics and the proposed ODRS.