Learning latent representations across multiple data domains using Lifelong VAEGAN

Abstract

The problem of catastrophic forgetting occurs in deep learning models trained on multiple databases in a sequential manner. Recently, generative replay mechanisms (GRM), have been proposed to reproduce previously learned knowledge aiming to reduce the forgetting. However, such approaches lack an appropriate inference model and therefore can not provide latent representations of data. In this paper, we propose a novel lifelong learning approach, namely the Lifelong VAEGAN (L-VAEGAN), which not only induces a powerful generative replay network but also learns meaningful latent representations, benefiting representation learning. L-VAEGAN can allow to automatically embed the information associated with different domains into several clusters in the latent space, while also capturing semantically meaningful shared latent variables, across different data domains. The proposed model supports many downstream tasks that traditional generative replay methods can not, including interpolation and inference across different data domains.

Accepted as a conference paper at ECCV 2020

References in corpus (8)

• Distilling the Knowledge in a Neural Network
• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• On Tiny Episodic Memories in Continual Learning
• Less-forgetting Learning in Deep Neural Networks
• Continual Unsupervised Representation Learning
• Adversarial Symmetric Variational Autoencoder
• Learning Discrete and Continuous Factors of Data via Alternating Disentanglement
• Symmetric Variational Autoencoder and Connections to Adversarial Learning