RMM: Reinforced Memory Management for Class-Incremental Learning

by   Yaoyao Liu, et al.

Class-Incremental Learning (CIL) [40] trains classifiers under a strict memory budget: in each incremental phase, learning is done for new data, most of which is abandoned to free space for the next phase. The preserved data are exemplars used for replaying. However, existing methods use a static and ad hoc strategy for memory allocation, which is often sub-optimal. In this work, we propose a dynamic memory management strategy that is optimized for the incremental phases and different object classes. We call our method reinforced memory management (RMM), leveraging reinforcement learning. RMM training is not naturally compatible with CIL as the past, and future data are strictly non-accessible during the incremental phases. We solve this by training the policy function of RMM on pseudo CIL tasks, e.g., the tasks built on the data of the 0-th phase, and then applying it to target tasks. RMM propagates two levels of actions: Level-1 determines how to split the memory between old and new classes, and Level-2 allocates memory for each specific class. In essence, it is an optimizable and general method for memory management that can be used in any replaying-based CIL method. For evaluation, we plug RMM into two top-performing baselines (LUCIR+AANets and POD+AANets [30]) and conduct experiments on three benchmarks (CIFAR-100, ImageNet-Subset, and ImageNet-Full). Our results show clear improvements, e.g., boosting POD+AANets by 3.6 respectively.


page 1

page 2

page 3

page 4


A Memory Transformer Network for Incremental Learning

We study class-incremental learning, a training setup in which new class...

Class-Incremental Exemplar Compression for Class-Incremental Learning

Exemplar-based class-incremental learning (CIL) finetunes the model with...

Remind of the Past: Incremental Learning with Analogical Prompts

Although data-free incremental learning methods are memory-friendly, acc...

Dataset Knowledge Transfer for Class-Incremental Learning without Memory

Incremental learning enables artificial agents to learn from sequential ...

Opacity of Memory Management in Software Transactional Memory

Opacity of Transactional Memory is proposed to be established by increme...

Meta-Aggregating Networks for Class-Incremental Learning

Class-Incremental Learning (CIL) aims to learn a classification model wi...

ScaIL: Classifier Weights Scaling for Class Incremental Learning

Incremental learning is useful if an AI agent needs to integrate data fr...

Please sign up or login with your details

Forgot password? Click here to reset