发表机构
Shanghai Jiao Tong University; Harbin Institute of Technology（上海交通大学; 哈尔滨工业大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
RestoreMore框架将预训练图像恢复模型作为冻结锚点，通过双层路由机制学习残差扩展模块，在不遗忘旧能力的前提下持续获取新恢复能力，经多基准实验验证有效。
AI 中文摘要
图像恢复模型通常在固定的能力集合上进行训练。当出现新的恢复需求时，现有解决方案通常会训练额外的模型，或者用新数据和历史数据联合重新训练原始模型。我们没有设计另一个恢复骨干，而是研究如何让已训练的恢复器在不遗忘先前学习能力的情况下持续获取新能力。我们提出了RestoreMore，这是一个持续能力扩展框架，它将预训练的恢复模型作为冻结的能力锚点，并为新出现的退化情况学习残差扩展模块。RestoreMore在多个特征阶段引入了面向能力的双层路由机制：第一个路由层级识别与当前输入相关的恢复能力，第二个层级选择并组合一组稀疏的互补退化专家。该设计使新引入的任务能够选择性地复用历史恢复知识，并逐步丰富可用于后续恢复任务的专家库。在广泛的恢复基准上进行的大量实验表明，RestoreMore在持续获取新恢复能力的同时，能够保留并提升先前学习到的能力。
英文摘要
Image restoration models are typically trained with a fixed set of capabilities. When new restoration requirements emerge, existing solutions usually train additional models or jointly retrain the original model with both new and historical data. Instead of designing another restoration backbone, we investigate how a trained restorer can continually acquire new capabilities without forgetting those learned previously. We propose RestoreMore, a continual capability-expansion framework that preserves the pretrained restoration model as a frozen capability anchor and learns residual expansion modules for newly arriving degradations. RestoreMore introduces a capability-oriented bi-level routing mechanism at multiple feature stages. The first routing level identifies restoration capabilities relevant to the current input, while the second selects and combines a sparse set of complementary degradation experts. This design enables newly introduced tasks to selectively reuse historical restoration knowledge and progressively enriches the expert bank available for subsequent restoration tasks. Extensive experiments on a wide range of restoration benchmarks demonstrate that RestoreMore consistently acquires new restoration abilities while preserving and improving previously learned capabilities.
