基于生成对抗元模型的面向回忆的持续学习
Recall-Oriented Continual Learning with Generative Adversarial Meta-Model
AI总结：
针对持续学习的稳定性-可塑性困境，提出受脑机制启发的双层面向回忆持续学习框架及生成对抗元模型GAMM，通过增量学习任务参数而非样本，在两类场景下兼顾新知识学习与旧知识稳定性。
摘要：
稳定性-可塑性困境是持续学习领域的一大核心挑战，其难点在于平衡两个相互冲突的目标：在学习新任务的同时维持对过往任务的性能表现。本文提出了面向回忆的持续学习框架以应对这一挑战，该框架受人类大脑分离稳定性与可塑性调控机制的能力启发，采用双层架构：推理网络负责高效获取新知识，生成网络则在必要时回忆过往知识。特别地，为最大化过往知识的稳定性，本文研究了不同表征下的知识复杂度，进而提出生成对抗元模型（GAMM），该模型以增量方式学习任务专属参数，而非学习任务的输入数据样本。实验结果表明，本文框架不仅能在不受任何干扰的情况下有效学习新知识，还能在任务感知和任务无关两种学习场景下实现过往知识的高稳定性。本文代码已公开于：https://github.com/bigdata-inha/recall-oriented-cl-framework。
英文摘要：
The stability-plasticity dilemma is a major challenge in continual learning, as it involves balancing the conflicting objectives of maintaining performance on previous tasks while learning new tasks. In this paper, we propose the recall-oriented continual learning framework to address this challenge. Inspired by the human brain's ability to separate the mechanisms responsible for stability and plasticity, our framework consists of a two-level architecture where an inference network effectively acquires new knowledge and a generative network recalls past knowledge when necessary. In particular, to maximize the stability of past knowledge, we investigate the complexity of knowledge depending on different representations, and thereby introducing generative adversarial meta-model (GAMM) that incrementally learns task-specific parameters instead of input data samples of the task. Through our experiments, we show that our framework not only effectively learns new knowledge without any disruption but also achieves high stability of previous knowledge in both task-aware and task-agnostic learning scenarios. Our code is available at: https://github.com/bigdata-inha/recall-oriented-cl-framework.
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。
