Recent Advances of Continual Learning in Computer Vision: An Overview

Abstract

In contrast to batch learning where all training data is available at once, continual learning represents a family of methods that accumulate knowledge and learn continuously with data available in sequential order. Similar to the human learning process with the ability of learning, fusing, and accumulating new knowledge coming at different time steps, continual learning is considered to have high practical significance. Hence, continual learning has been studied in various artificial intelligence tasks. In this paper, we present a comprehensive review of the recent progress of continual learning in computer vision. In particular, the works are grouped by their representative techniques, including regularization, knowledge distillation, memory, generative replay, parameter isolation, and a combination of the above techniques. For each category of these techniques, both its characteristics and applications in computer vision are presented. At the end of this overview, several subareas, where continuous knowledge accumulation is potentially helpful while continual learning has not been well studied, are discussed.

June 2024 Version

References in corpus (29)

• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Conditional Image Synthesis With Auxiliary Classifier GANs
• Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigence
• Continual Learning Through Synaptic Intelligence
• Accurate sampling using Langevin dynamics
• Optimizing Neural Networks with Kronecker-factored Approximate Curvature
• Reinforced Continual Learning
• Generative replay with feedback connections as a general strategy for continual learning
• Online Structured Laplace Approximations For Overcoming Catastrophic Forgetting
• Incremental Learning for Semantic Segmentation of Large-Scale Remote Sensing Data
• A Neural Dirichlet Process Mixture Model for Task-Free Continual Learning
• Understanding the Role of Training Regimes in Continual Learning
• Coresets via Bilevel Optimization for Continual Learning and Streaming
• Gradient based sample selection for online continual learning
• Learning to Continually Learn
• CPR: Classifier-Projection Regularization for Continual Learning
• Incremental Few-Shot Learning with Attention Attractor Networks
• Efficient Continual Learning with Modular Networks and Task-Driven Priors
• Continual Learning with Node-Importance based Adaptive Group Sparse Regularization
• Beyond Shared Hierarchies: Deep Multitask Learning through Soft Layer Ordering
• Wandering Within a World: Online Contextualized Few-Shot Learning
• Reconciling meta-learning and continual learning with online mixtures of tasks
• ContCap: A scalable framework for continual image captioning
• Deep Networks from the Principle of Rate Reduction
• Bayesian Structure Adaptation for Continual Learning
• Optimizing Reusable Knowledge for Continual Learning via Metalearning
• Graph-Based Continual Learning
• Kernel Continual Learning

Cited by in corpus (4)

• Incremental Few-Shot Semantic Segmentation via Embedding Adaptive-Update and Hyper-class Representation
• Multimodal Parameter-Efficient Few-Shot Class Incremental Learning
• Inherit with Distillation and Evolve with Contrast: Exploring Class Incremental Semantic Segmentation Without Exemplar Memory
• ECG-CL: A Comprehensive Electrocardiogram Interpretation Method Based on Continual Learning