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

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

30 Oct 2021NeurIPS 2021 12arXiv:2111.01549archive 2025-07-28

Guangyuan Shi, Jiaxin Chen, Wenlong Zhang, Li-Ming Zhan, Xiao-Ming Wu

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.

Paper PDF Conference PDF Code

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

moukamisama/f2m officialmentioned in paperpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Few-Shot Class-Incremental Learning Few-Shot Learning Incremental Learning

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Few-Shot Class-Incremental Learning	CIFAR-100	F2M	Average Accuracy	53.69	#10 of 11	Archive leaderboard	report
Few-Shot Class-Incremental Learning	CIFAR-100	F2M	Last Accuracy	44.65	#10 of 11	Archive leaderboard	report
Few-Shot Class-Incremental Learning	mini-Imagenet	F2M	Average Accuracy	54.89	#9 of 12	Archive leaderboard	report
Few-Shot Class-Incremental Learning	mini-Imagenet	F2M	Last Accuracy	47.84	#9 of 12	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections