LibContinual：面向真实连续学习的综合性库
LibContinual: A Comprehensive Library towards Realistic Continual Learning
AI总结：
LibContinual提出一个综合性库，旨在解决持续学习中的现实约束问题，通过统一框架和严格评估揭示方法的性能下降。
摘要：
持续学习（CL）面临的核心挑战是灾难性遗忘，即适应新任务会损害对先前任务的性能。尽管该领域已发展出多样化的技术，但这些快速发展的方法导致了研究领域的碎片化。缺乏统一的框架，包括不一致的实现、冲突的依赖关系和不同的评估协议，使得公平比较和可重复研究变得更加困难。为了解决这一挑战，我们提出了LibContinual，一个全面且可重复的库，旨在作为真实连续学习的基础平台。该库基于高内聚、低耦合的模块化架构，整合了五种主要方法学类别中的19种代表性算法，提供标准化的执行环境。同时，利用这一统一框架，我们系统地识别并研究了主流评估中普遍存在的三个隐含假设：（1）离线数据可访问性，（2）不受监管的内存资源，（3）任务内语义同质性。我们认为这些假设往往高估了连续学习方法在现实世界中的适用性。通过我们的全面分析，使用严格的在线连续学习设置、一种新的统一内存预算协议和一个提议的类别随机化设置，我们揭示了在这些现实约束下，许多代表性连续学习方法的显著性能下降。我们的研究强调了资源感知和语义稳健的连续学习策略的必要性，并提供了LibContinual作为未来真实连续学习研究的基础工具包。源代码可在https://github.com/RL-VIG/LibContinual上获得。
英文摘要：
A fundamental challenge in Continual Learning (CL) is catastrophic forgetting, where adapting to new tasks degrades the performance on previous ones. While the field has evolved with diverse methods, this rapid surge in diverse methodologies has culminated in a fragmented research landscape. The lack of a unified framework, including inconsistent implementations, conflicting dependencies, and varying evaluation protocols, makes fair comparison and reproducible research increasingly difficult. To address this challenge, we propose LibContinual, a comprehensive and reproducible library designed to serve as a foundational platform for realistic CL. Built upon a high-cohesion, low-coupling modular architecture, LibContinual integrates 19 representative algorithms across five major methodological categories, providing a standardized execution environment. Meanwhile, leveraging this unified framework, we systematically identify and investigate three implicit assumptions prevalent in mainstream evaluation: (1) offline data accessibility, (2) unregulated memory resources, and (3) intra-task semantic homogeneity. We argue that these assumptions often overestimate the real-world applicability of CL methods. Through our comprehensive analysis using strict online CL settings, a novel unified memory budget protocol, and a proposed category-randomized setting, we reveal significant performance drops in many representative CL methods when subjected to these real-world constraints. Our study underscores the necessity of resource-aware and semantically robust CL strategies, and offers LibContinual as a foundational toolkit for future research in realistic continual learning. The source code is available from \href{https://github.com/RL-VIG/LibContinual}{https://github.com/RL-VIG/LibContinual}.
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。
