Online Prototype Learning for Online Continual Learning

By Yujie Wei, Jiaxin Ye, Zhizhong Huang, Junping Zhang, Hongming Shan. 2023

82 citations recorded by the source corpus.

Online continual learning (CL) studies the problem of learning continuously from a single-pass data stream while adapting to new data and mitigating catastrophic forgetting. Recently, by storing a small subset of old data, replay-based methods have shown promising performance. Unlike previous methods that focus on sample storage or knowledge distillation against catastrophic forgetting, this paper aims to understand why the online learning models fail to generalize well from a new perspective of shortcut learning. We identify shortcut learning as the key limiting factor for online CL, where the learned features may be biased, not generalizable to new tasks, and may have an adverse impact on knowledge distillation. To tackle this issue, we present the online prototype learning (OnPro) framework for online CL. First, we propose online prototype equilibrium to learn representative features against shortcut learning and discriminative features to avoid class confusion, ultimately achieving an equilibrium status that separates all seen classes well while learning new classes. Second, with the feedback of online prototypes, we devise a novel adaptive prototypical feedback mechanism to sense the classes that are easily misclassified and then enhance their boundaries. Extensive experimental results on widely-used benchmark datasets demonstrate the superior performance of OnPro over the state-of-the-art baseline methods. Source code is available at https://github.com/weilllllls/OnPro.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Improving Plasticity in Online Continual Learning via Collaborative Learning (2023, methodology): “Replay-based methods [6, 7, 10, 15, 21, 22, 34, 38, 48] maintain a small memory buffer that stores a few old training samples.”
Learning from One Continuous Video Stream (2023, background): “One problem that has been identified in this setting is cheating by exploiting labels in the data [16], or by learning spurious features [46] that have limited generalisation power.”
Orchestrate Latent Expertise: Advancing Online Continual Learning with Multi-Level Supervision and Reverse Self-Distillation (2024, methodology): “Following the setup of previous OCL research [21, 22, 67], we split CIFAR-100 into 10 disjoint tasks with 10 classes per task (500/100 training/testing samples per class) and split Tiny-ImageNet into 100 disjoint tasks with 2 classes per task (500/50 training/testing samples per class),…”