Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by Finding Flat Minima

2021/10/30 by Guangyuan Shi, Jiaxin Chen, Shi, Guangyuan +7 · 8 citations
Computer Science · Medicine · #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG) #Multimodal Machine Learning Applications

paper · pdf · doi:10.48550/arxiv.2111.01549

openalex publication_date 2021/10/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

This paper considers incremental few-shot learning, which requires a model to continually recognize new categories with only a few examples provided. Our study shows that existing methods severely suffer from catastrophic forgetting, a well-known problem in incremental learning, which is aggravated due to data scarcity and imbalance in the few-shot setting. Our analysis further suggests that to prevent catastrophic forgetting, actions need to be taken in the primitive stage -- the training of base classes instead of later few-shot learning sessions. Therefore, we propose to search for flat local minima of the base training objective function and then fine-tune the model parameters within the flat region on new tasks. In this way, the model can efficiently learn new classes while preserving the old ones. Comprehensive experimental results demonstrate that our approach outperforms all prior state-of-the-art methods and is very close to the approximate upper bound. The source code is available at https://github.com/moukamisama/F2M.

Cited by

• Hierarchical Semantic Tree Anchoring for CLIP-Based Class-Incremental Learning
• More Than Memory Savings: Zeroth-Order Optimization Mitigates Forgetting in Continual Learning
• SAFA-SNN: Sparsity-Aware On-Device Few-Shot Class-Incremental Learning with Fast-Adaptive Structure of Spiking Neural Network
• Hierarchical Representation Matching for CLIP-based Class-Incremental Learning
• MoTiC: Momentum Tightness and Contrast for Few-Shot Class-Incremental Learning
• Control Disturbance Rejection in Neural ODEs
• C-Flat++: Towards a More Efficient and Powerful Framework for Continual Learning
• Learning from Oblivion: Predicting Knowledge Overflowed Weights via Retrodiction of Forgetting

Related

• Few-Shot Class-Incremental Learning with Prior Knowledge
• Controllable Forgetting Mechanism for Few-Shot Class-Incremental Learning
• Latest Advancements Towards Catastrophic Forgetting under Data Scarcity: A Comprehensive Survey on Few-Shot Class Incremental Learning
• Few-Shot Continual Learning via Flat-to-Wide Approaches
• Generalized and Incremental Few-Shot Learning by Explicit Learning and Calibration without Forgetting