自适应原型回放用于类增量语义分割
Adaptive Prototype Replay for Class Incremental Semantic Segmentation
AI总结：
本文提出自适应原型回放（Adapter）方法，通过动态更新原型和不确定性约束损失，解决类增量语义分割中表示更新与固定原型不匹配的问题，在Pascal VOC和ADE20K上达到最先进性能。
摘要：
类增量语义分割（CISS）旨在连续学习步骤中分割新类别，同时防止旧知识遗忘。现有方法通过使用存储的原型或特征回放先前学习类别的分布来缓解灾难性遗忘。然而，它们忽略了一个关键问题：在CISS中，类别知识的表示通过增量学习不断更新，而原型回放方法保持固定的原型。更新后的表示与固定原型之间的不匹配限制了原型回放策略的有效性。为解决此问题，本文提出用于CISS的自适应原型回放（Adapter）。Adapter包含自适应偏差补偿（ADC）策略和不确定性感知约束（UAC）损失。具体而言，ADC策略基于估计的表示偏移距离动态更新存储的原型，以匹配旧类别更新后的表示。UAC损失减少预测不确定性，聚合判别性特征以帮助生成紧凑原型。此外，我们引入基于补偿的原型相似性判别（CPD）损失，以确保相似原型之间有足够的区分度，从而增强自适应原型回放策略的效率。在Pascal VOC和ADE20K数据集上的大量实验表明，Adapter取得了最先进的结果，并在各种CISS任务中证明有效，特别是在具有挑战性的多步场景中。代码和模型可在https://github.com/zhu-gl-ux/Adapter获取。
英文摘要：
Class incremental semantic segmentation (CISS) aims to segment new classes during continual steps while preventing the forgetting of old knowledge. Existing methods alleviate catastrophic forgetting by replaying distributions of previously learned classes using stored prototypes or features. However, they overlook a critical issue: in CISS, the representation of class knowledge is updated continuously through incremental learning, whereas prototype replay methods maintain fixed prototypes. This mismatch between updated representation and fixed prototypes limits the effectiveness of the prototype replay strategy. To address this issue, we propose the Adaptive prototype replay (Adapter) for CISS in this paper. Adapter comprises an adaptive deviation compen sation (ADC) strategy and an uncertainty-aware constraint (UAC) loss. Specifically, the ADC strategy dynamically updates the stored prototypes based on the estimated representation shift distance to match the updated representation of old class. The UAC loss reduces prediction uncertainty, aggregating discriminative features to aid in generating compact prototypes. Additionally, we introduce a compensation-based prototype similarity discriminative (CPD) loss to ensure adequate differentiation between similar prototypes, thereby enhancing the efficiency of the adaptive prototype replay strategy. Extensive experiments on Pascal VOC and ADE20K datasets demonstrate that Adapter achieves state-of-the-art results and proves effective across various CISS tasks, particularly in challenging multi-step scenarios. The code and model is available at https://github.com/zhu-gl-ux/Adapter.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
