Journal
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
Volume 45, Issue 5, Pages 5513-5533Publisher
IEEE COMPUTER SOC
DOI: 10.1109/TPAMI.2022.3213473
Keywords
Task analysis; Training; Network architecture; Learning systems; Image classification; Training data; Privacy; Class-incremental learning; continual learning; incremental learning; lifelong learning; catastrophic forgetting
Ask authors/readers for more resources
For future learning systems, incremental learning is desirable due to its efficient resource usage, reduced memory usage, and resemblance to human learning. The main challenge for incremental learning is catastrophic forgetting. This paper provides a comprehensive survey of existing class-incremental learning methods for image classification and performs extensive experimental evaluations on thirteen methods.
For future learning systems, incremental learning is desirable because it allows for: efficient resource usage by eliminating the need to retrain from scratch at the arrival of new data; reduced memory usage by preventing or limiting the amount of data required to be stored - also important when privacy limitations are imposed; and learning that more closely resembles human learning. The main challenge for incremental learning is catastrophic forgetting, which refers to the precipitous drop in performance on previously learned tasks after learning a new one. Incremental learning of deep neural networks has seen explosive growth in recent years. Initial work focused on task-incremental learning, where a task-ID is provided at inference time. Recently, we have seen a shift towards class-incremental learning where the learner must discriminate at inference time between all classes seen in previous tasks without recourse to a task-ID. In this paper, we provide a complete survey of existing class-incremental learning methods for image classification, and in particular, we perform an extensive experimental evaluation on thirteen class-incremental methods. We consider several new experimental scenarios, including a comparison of class-incremental methods on multiple large-scale image classification datasets, an investigation into small and large domain shifts, and a comparison of various network architectures.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available