Mnemonics training: Multi-class incremental learning without forgetting

• Main Authors: LIU, Yaoyao, SU, Yuting, LIU, An-An, SCHIELE, Bernt, SUN, Qianru
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2020
• Subjects: Databases and Information Systems; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/sis_research/5593; https://ink.library.smu.edu.sg/context/sis_research/article/6596/viewcontent/Liu_Mnemonics_Training_Multi_Class_Incremental_Learning_Without_Forgetting_CVPR_2020_paper.pdf
• Institution: Singapore Management University

Multi-Class Incremental Learning (MCIL) aims to learn new concepts by incrementally updating a model trained on previous concepts. However, there is an inherent trade-off to effectively learning new concepts without catastrophic forgetting of previous ones. To alleviate this issue, it has been proposed to keep around a few examples of the previous concepts but the effectiveness of this approach heavily depends on the representativeness of these examples. This paper proposes a novel and automatic framework we call mnemonics, where we parameterize exemplars and make them optimizable in an end-to-end manner. We train the framework through bilevel optimizations, i.e., model-level and exemplar-level. We conduct extensive experiments on three MCIL benchmarks, CIFAR-100, ImageNet-Subset and ImageNet, and show that using mnemonics exemplars can surpass the state-of-the-art by a large margin. Interestingly and quite intriguingly, the mnemonics exemplars tend to be on the boundaries between different classes.