融合专用与通用适配器，提升预训练模型持续学习性能
Integrating Task-Specific and Universal Adapters for Pre-Trained Model-based Class-Incremental Learning
- 设计专用与通用双适配器架构，分别捕捉任务特性和共享知识
- 通过熵值选择机制动态选适配器，避免错误推理
- 适用于需要持续学习新类别的实际场景，如图像分类
类别增量学习（CIL）要求系统在不遗忘旧类的前提下持续学习新类别。现有基于预训练模型的CIL方法通常冻结预训练网络，使用轻量级模块如适配器进行增量适应。然而，推理时模块选择不当会降低性能，且专用适配器常忽略跨任务共享的通用知识，导致相似类别区分错误。本文提出集成专用与通用适配器（TUNA）的方法：训练专用适配器以提取各任务关键特征，并引入基于熵的适配器选择机制；同时采用适配器融合策略构建一个编码跨任务判别性特征的通用适配器。推理时融合两类适配器预测结果，兼顾专精与通用知识。在多个基准数据集上的大量实验表明，该方法达到当前最优性能。代码已开源。
原文摘要 · Abstract (English)
Class-Incremental Learning (CIL) requires a learning system to continually learn new classes without forgetting. Existing pre-trained model-based CIL methods often freeze the pre-trained network and adapt to incremental tasks using additional lightweight modules such as adapters. However, incorrect module selection during inference hurts performance, and task-specific modules often overlook shared general knowledge, leading to errors on distinguishing between similar classes across tasks. To address the aforementioned challenges, we propose integrating Task-Specific and Universal Adapters (TUNA) in this paper. Specifically, we train task-specific adapters to capture the most crucial features relevant to their respective tasks and introduce an entropy-based selection mechanism to choose the most suitable adapter. Furthermore, we leverage an adapter fusion strategy to construct a universal adapter, which encodes the most discriminative features shared across tasks. We combine task-specific and universal adapter predictions to harness both specialized and general knowledge during inference. Extensive experiments on various benchmark datasets demonstrate the state-of-the-art performance of our approach. Code is available at: https://github.com/LAMDA-CL/ICCV2025-TUNA
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。