Search papers, labs, and topics across Lattice.
This paper introduces Loop-Mamba, a novel state-space framework designed for restoring old photographs by addressing multiple coupled degradations. The method employs a Semantic-Guided Degradation Estimator (SGDE) to predict local and global degradation metrics, allowing for informed state evolution, while a Shared Structural Memory Mamba (S虏M-Mamba) facilitates persistent state propagation across iterations. Experimental results on the SynOld benchmark show that Loop-Mamba significantly outperforms existing restoration techniques, achieving superior results in both traditional metrics and the newly proposed Old Photo Damage Recovery Score (ODRS).
Loop-Mamba not only restores old photos more effectively but also introduces a novel degradation-aware framework that outperforms state-of-the-art methods by leveraging persistent memory and semantic guidance.
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.
