分段存储记忆，让模型像人一样持续学新类不遗忘旧知识。
Partitioned Memory Storage Inspired Few-Shot Class-Incremental learning
- 为每个学习阶段训练独立模型，避免灾难性遗忘。
- 在CIFAR-100和mini-ImageNet上达到当前最优性能。
- 结合不确定性量化实现测试时智能模型选择。
当前主流深度学习技术高度依赖大量训练数据，且难以适应动态环境，与人类智能存在显著差距。为弥合这一鸿沟，少样本类别增量学习（FSCIL）应运而生，旨在用少量样本持续学习新类别，同时不遗忘旧知识。现有FSCIL方法通常使用单一模型跨会话学习，不可避免地陷入稳定性与可塑性之间的矛盾。不同于机器，人类将不同类型的知识存储在不同脑皮层中。受此启发，本文提出为每个学习会话训练独立模型，从而天然防止灾难性遗忘。测试阶段，引入不确定性量化（UQ）进行模型部署决策。该方法为FSCIL提供了新视角，并在CIFAR-100和mini-ImageNet数据集上取得当前最优表现。
原文摘要 · Abstract (English)
Current mainstream deep learning techniques exhibit an over-reliance on extensive training data and a lack of adaptability to the dynamic world, marking a considerable disparity from human intelligence. To bridge this gap, Few-Shot Class-Incremental Learning (FSCIL) has emerged, focusing on continuous learning of new categories with limited samples without forgetting old knowledge. Existing FSCIL studies typically use a single model to learn knowledge across all sessions, inevitably leading to the stability-plasticity dilemma. Unlike machines, humans store varied knowledge in different cerebral cortices. Inspired by this characteristic, our paper aims to develop a method that learns independent models for each session. It can inherently prevent catastrophic forgetting. During the testing stage, our method integrates Uncertainty Quantification (UQ) for model deployment. Our method provides a fresh viewpoint for FSCIL and demonstrates the state-of-the-art performance on CIFAR-100 and mini-ImageNet datasets.
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。