Logo image
Unsupervised Generative Variational Continual Learning
Conference proceeding

Unsupervised Generative Variational Continual Learning

Liu Guimeng, Guo Yang, Cheryl Wong Sze Yin, Ponnuthurai Nagartnam Suganathan, Ramasamy Savitha, IEEE and Guimeng Liu
Proceedings - International Conference on Image Processing, pp.4028-4032
16/10/2022

Abstract

Adaptation models Benchmark testing Continual Learning Image coding Neurons Task analysis Training Uncertainty Unsupervised Variational Inference
Continual learning aims at learning a sequence of tasks without forgetting any task. While most of the existing literature in continual learning is aimed at class incremental learning in a supervised setting, there is an enormous potential for unsupervised continual learning using generative models. This paper proposes a combination of architectural pruning and neuron addition in generative variational models toward unsupervised generative continual learning (UGCL). Evaluations on standard benchmark data sets demonstrate the superior generative ability of the proposed method.

Metrics

1 Record Views

Details

Logo image