Compacting, Picking and Growing for Unforgetting Continual Learning

TLDR

Continual lifelong learning is essential to many applications. The paper proposes a simple yet effective approach for continual deep learning. Our approach leverages deep model compression, critical weight selection, and progressive network expansion, integrating them iteratively to create a scalable incremental learning method. The method is easy to implement, prevents forgetting, maintains model compactness while allowing expansion, and outperforms independent task training by leveraging accumulated knowledge, as shown by experimental results.

Abstract

Continual lifelong learning is essential to many applications. In this paper, we propose a simple but effective approach to continual deep learning. Our approach leverages the principles of deep model compression, critical weights selection, and progressive networks expansion. By enforcing their integration in an iterative manner, we introduce an incremental learning method that is scalable to the number of sequential tasks in a continual learning process. Our approach is easy to implement and owns several favorable characteristics. First, it can avoid forgetting (i.e., learn new tasks while remembering all previous tasks). Second, it allows model expansion but can maintain the model compactness when handling sequential tasks. Besides, through our compaction and selection/expansion mechanism, we show that the knowledge accumulated through learning previous tasks is helpful to build a better model for the new tasks compared to training the models independently with tasks. Experimental results show that our approach can incrementally learn a deep model tackling multiple tasks without forgetting, while the model compactness is maintained with the performance more satisfiable than individual task training.