Active Continual Learning: On Balancing Knowledge Retention and Learnability

TL;DR

This paper explores active continual learning, focusing on balancing knowledge retention and rapid learnability across different scenarios, and investigates how various algorithms perform in this context.

Contribution

It introduces the problem of active continual learning with unlabelled data pools and annotation budgets, analyzing the interplay of AL and CL algorithms across scenarios.

Findings

Conditioning AL query strategies on previous annotations improves performance in domain and task-incremental learning.

A gap exists in balancing forgetting and learning in class-incremental scenarios.

Experiments reveal trade-offs between knowledge retention and learnability in ACL.

Abstract

Acquiring new knowledge without forgetting what has been learned in a sequence of tasks is the central focus of continual learning (CL). While tasks arrive sequentially, the training data are often prepared and annotated independently, leading to the CL of incoming supervised learning tasks. This paper considers the under-explored problem of active continual learning (ACL) for a sequence of active learning (AL) tasks, where each incoming task includes a pool of unlabelled data and an annotation budget. We investigate the effectiveness and interplay between several AL and CL algorithms in the domain, class and task-incremental scenarios. Our experiments reveal the trade-off between two contrasting goals of not forgetting the old knowledge and the ability to quickly learn new knowledge in CL and AL, respectively. While conditioning the AL query strategy on the annotations collected for the previous tasks leads to improved task performance on the domain and task incremental learning, our proposed forgetting-learning profile suggests a gap in balancing the effect of AL and CL for the class-incremental scenario.