AI 中文总结
针对开放词汇变化检测现有方法的缺陷，提出认知记忆引导的CogVis框架，通过感知-记忆-验证范式解耦相关任务，在七个基准上达最优性能，推理吞吐量提升28.50%。
AI 中文摘要
地表监测需要能够识别任意语义类别的变化检测模型，开放词汇变化检测（Open-Vocabulary Change Detection, OVCD）满足这一需求。然而现有方法常将时间感知、语义判别与区域验证纠缠在一起，导致结果不稳定且计算冗余。受人类视觉变化感知启发，我们提出CogVis，一种认知记忆引导的框架，将OVCD重新表述为感知-记忆-验证范式。CogVis首先采用场景变化感知器（Scene Change Perceptron, SCP）从冻结的双时相特征中提取可复用、与类别无关的变化先验，从而将时间证据与语义类别决策解耦。随后，语义记忆校准器（Semantic Memory Calibrator, SMC）通过动态估计特定于图像-查询的决策阈值，补偿与类别相关的分数偏移。最后，自适应区域过滤器（Adaptive Region Filter, ARF）利用学习到的语义、时间和结构可靠性过滤连通候选区域。在涵盖语义变化检测、二值变化定位和建筑损伤评估的七个基准上的实验表明，CogVis在所有评估数据集上均达到了最优性能；通过共享场景级别的变化感知，CogVis进一步避免了在各查询间重复进行与类别无关的时间感知，推理吞吐量提升了28.50%。
英文摘要
Earth-surface monitoring requires change detection models capable of recognizing arbitrary semantic categories. Open-Vocabulary Change Detection (OVCD) addresses this need. However, existing methods often entangle temporal perception, semantic discrimination, and region verification, causing unstable results and redundant computation. Inspired by human visual change perception, we propose CogVis, a cognitive memory-guided framework that reformulates OVCD as a perception-memory-verification paradigm. CogVis first employs a Scene Change Perceptron (SCP) to extract a reusable, category-agnostic change prior from frozen bi-temporal features, thereby decoupling temporal evidence from semantic category decisions. A Semantic Memory Calibrator (SMC) then compensates for category-dependent score shifts by dynamically estimating an image-query-specific decision threshold. Finally, an Adaptive Region Filter (ARF) filters connected candidates using learned semantic, temporal, and structural reliability. Experiments on seven benchmarks spanning semantic change detection, binary change localization, and building-damage assessment show that CogVis achieves state-of-the-art performance across all evaluated datasets. By sharing scene-level change perception, CogVis further avoids repeating category-agnostic temporal perception across queries and improves inference throughput by 28.50%.
Comments19 pages, 11 figures, including 3 supplementary figures. Code: https://github.com/KotlinWang/CogVis
