A continual learning survey: Defying forgetting in classification tasks

TL;DR

This survey reviews and compares state-of-the-art continual learning methods for classification, proposing a new framework to balance stability and plasticity, and empirically evaluates their performance on multiple benchmarks.

Contribution

It provides a comprehensive taxonomy, introduces a novel stability-plasticity framework, and empirically compares 11 methods across diverse datasets.

Findings

Different methods vary in memory and computation requirements.

Model capacity and regularization influence continual learning performance.

Task order impacts the effectiveness of continual learning methods.

Abstract

Artificial neural networks thrive in solving the classification problem for a particular rigid task, acquiring knowledge through generalized learning behaviour from a distinct training phase. The resulting network resembles a static entity of knowledge, with endeavours to extend this knowledge without targeting the original task resulting in a catastrophic forgetting. Continual learning shifts this paradigm towards networks that can continually accumulate knowledge over different tasks without the need to retrain from scratch. We focus on task incremental classification, where tasks arrive sequentially and are delineated by clear boundaries. Our main contributions concern 1) a taxonomy and extensive overview of the state-of-the-art, 2) a novel framework to continually determine the stability-plasticity trade-off of the continual learner, 3) a comprehensive experimental comparison of 11 state-of-the-art continual learning methods and 4 baselines. We empirically scrutinize method strengths and weaknesses on three benchmarks, considering Tiny Imagenet and large-scale unbalanced iNaturalist and a sequence of recognition datasets. We study the influence of model capacity, weight decay and dropout regularization, and the order in which the tasks are presented, and qualitatively compare methods in terms of required memory, computation time, and storage.