Deep Generative Dual Memory Network for Continual Learning

2017/10/28 by Nitin Kamra, Umang Gupta, Kamra, Nitin +3 · 3 citations
Computer Science · Neuroscience · #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG) #Multimodal Machine Learning Applications #Sleep and Wakefulness Research

paper · pdf · doi:10.48550/arxiv.1710.10368

openalex publication_date 2017/10/28 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Despite advances in deep learning, neural networks can only learn multiple tasks when trained on them jointly. When tasks arrive sequentially, they lose performance on previously learnt tasks. This phenomenon called catastrophic forgetting is a fundamental challenge to overcome before neural networks can learn continually from incoming data. In this work, we derive inspiration from human memory to develop an architecture capable of learning continuously from sequentially incoming tasks, while averting catastrophic forgetting. Specifically, our contributions are: (i) a dual memory architecture emulating the complementary learning systems (hippocampus and the neocortex) in the human brain, (ii) memory consolidation via generative replay of past experiences, (iii) demonstrating advantages of generative replay and dual memories via experiments, and (iv) improved performance retention on challenging tasks even for low capacity models. Our architecture displays many characteristics of the mammalian memory and provides insights on the connection between sleep and learning.

Citations

• Auto-Encoding Variational Bayes
• Gradient-based learning applied to document recognition
• An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks
• Catastrophic Forgetting, Rehearsal and Pseudorehearsal
• Adversarial Feature Learning
• Learning without Forgetting
• PathNet: Evolution Channels Gradient Descent in Super Neural Networks
• Continual Learning with Deep Generative Replay
• Gradient Episodic Memory for Continual Learning
• Training Gaussian Mixture Models at Scale via Coresets
• Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks

Cited by

• Dynamic Dual Buffer with Divide-and-Conquer Strategy for Online Continual Learning
• OpenLORIS-Object: A Robotic Vision Dataset and Benchmark for Lifelong Deep Learning
• Self-Supervised Learning Aided Class-Incremental Lifelong Learning

Related

• Continual Learning with Deep Generative Replay
• Self-recovery of memory via generative replay
• Semi-parametric Memory Consolidation: Towards Brain-like Deep Continual Learning
• Learning Fast, Learning Slow: A General Continual Learning Method based on Complementary Learning System
• Teaching AI to Remember: Insights from Brain-Inspired Replay in Continual Learning