Abstract:Modern artificial intelligence systems require class-incremental learning while suffering from catastrophic forgetting in many real-world applications. Due to the missing knowledge of past data, performance substantially degrades. Recent methods often used knowledge distillation and bias correction to avoid catastrophic forgetting caused by cognitive bias. However, since these methods mainly learn all samples indiscriminately, the model is hard to learn what it truly needs from the data stream to balance the new and old knowledge, leading to inevitable forgetting. Instead of considering each sample indiscriminately, the model should learn from its curious samples automatically. To tackle this problem, we propose a curiosity-driven class-incremental learning approach via adaptive sample selection for learning a more generalized model with fewer ineffective updates. Specifically, our method quantifies the model’s curiosity in each sample by two properties: uncertainty and novelty. Our model learns informative samples selectively during training utilizing the proposed uncertainty property, which benefits the classification decision boundary. In light of the imbalanced data, a novelty property is used to selectively optimize the model by employing dissimilar samples, endowing it with more robustness and less cognitive bias. Our method successfully reduces catastrophic forgetting and can be flexibly incorporated with other techniques. Extensive experiments and in-depth analysis on the CIFAR-100, Tiny-ImageNet and Caltech-101 datasets show that our approach outperforms competing methods for class-incremental learning in terms of preventing catastrophic forgetting.

A Strategy for an Uncompromising Incremental Learner

Incremental Learning Using Conditional Adversarial Networks

An Efficient Strategy for Catastrophic Forgetting Reduction in Incremental Learning

Large Scale Incremental Learning

Class incremental learning with probability dampening and cascaded gated classifier

Few-Shot Incremental Learning for Label-to-Image Translation

Incremental Learning of Object Detectors without Catastrophic Forgetting

Incremental Learning In Online Scenario

Incremental Classifier Learning with Generative Adversarial Networks

Curiosity-Driven Class-Incremental Learning Via Adaptive Sample Selection

Class-Incremental Learning via Knowledge Amalgamation

Model Behavior Preserving for Class-Incremental Learning

Prototypes Sampling Mechanism for Class Incremental Learning

Neural Collapse Terminus: A Unified Solution for Class Incremental Learning and Its Variants

Incremental Concept Learning via Online Generative Memory Recall

Incremental Learning with Maximum Entropy Regularization: Rethinking Forgetting and Intransigence

Incremental Learning Via Rate Reduction

Learning a Unified Classifier Incrementally via Rebalancing

Overcoming Catastrophic Forgetting with Unlabeled Data in the Wild

Long-Tail Class Incremental Learning Via Independent Sub-prototype Construction

Rethinking Class-Incremental Learning from a Dynamic Imbalanced Learning Perspective