Abstract:Large deep learning models are impressive, but they struggle when real-time data is not available. Few-shot class-incremental learning (FSCIL) poses a significant challenge for deep neural networks to learn new tasks from just a few labeled samples without forgetting the previously learned ones. This setup can easily leads to catastrophic forgetting and overfitting problems, severely affecting model performance. Studying FSCIL helps overcome deep learning model limitations on data volume and acquisition time, while improving practicality and adaptability of machine learning models. This paper provides a comprehensive survey on FSCIL. Unlike previous surveys, we aim to synthesize few-shot learning and incremental learning, focusing on introducing FSCIL from two perspectives, while reviewing over 30 theoretical research studies and more than 20 applied research studies. From the theoretical perspective, we provide a novel categorization approach that divides the field into five subcategories, including traditional machine learning methods, meta learning-based methods, feature and feature space-based methods, replay-based methods, and dynamic network structure-based methods. We also evaluate the performance of recent theoretical research on benchmark datasets of FSCIL. From the application perspective, FSCIL has achieved impressive achievements in various fields of computer vision such as image classification, object detection, and image segmentation, as well as in natural language processing and graph. We summarize the important applications. Finally, we point out potential future research directions, including applications, problem setups, and theory development. Overall, this paper offers a comprehensive analysis of the latest advances in FSCIL from a methodological, performance, and application perspective.

Mine-Distill-Prototypes for Complete Few-Shot Class-Incremental Learning in Image Classification

Few-Shot Class-Incremental Learning Via Class-Aware Bilateral Distillation

Pseudo Initialization Based Few-Shot Class Incremental Learning

Few Shot Class Incremental Learning via Efficient Prototype Replay and Calibration

A Cognition-Driven Framework for Few-Shot Class-Incremental Learning

Few-Shot Class-Incremental Learning via Training-Free Prototype Calibration

Uncertainty-Guided Semi-Supervised Few-Shot Class-Incremental Learning With Knowledge Distillation

On Distilling the Displacement Knowledge for Few-Shot Class-Incremental Learning

Few-shot class-incremental learning based on representation enhancement

Coarse-To-Fine Incremental Few-Shot Learning

Synthesized Feature based Few-Shot Class-Incremental Learning on a Mixture of Subspaces

Self-supervised Contrastive Feature Refinement for Few-Shot Class-Incremental Learning.

Few-Shot Incremental Learning with Continually Evolved Classifiers

A survey on few-shot class-incremental learning

M2SD:Multiple Mixing Self-Distillation for Few-Shot Class-Incremental Learning

Few-Shot Class-Incremental Learning by Sampling Multi-Phase Tasks

Analogical Learning-Based Few-Shot Class-Incremental Learning

Compositional Few-Shot Class-Incremental Learning

Dynamic Support Network for Few-Shot Class Incremental Learning

FineFMPL: Fine-grained Feature Mining Prompt Learning for Few-Shot Class Incremental Learning

Learning Primitive-aware Discriminative Representations for Few-shot Learning