Generative Models from the perspective of Continual Learning

Timothée Lesort, Hugo Caselles-Dupré, Michael Garcia-Ortiz, Jean-François Goudou, David Filliat

Abstract: Which generative model is the most suitable for Continual Learning? This paper aims at evaluating and comparing generative models on disjoint sequential image generation tasks. We investigate how several models learn and forget, considering various strategies: rehearsal, regularization, generative replay and fine-tuning. We used two quantitative metrics to estimate the generation quality and memory ability. We experiment with sequential tasks on three commonly used benchmarks for Continual Learning (MNIST, Fashion MNIST and CIFAR10). We found that among all models, the original GAN performs best and among Continual Learning strategies, generative replay outperforms all other methods. Even if we found satisfactory combinations on MNIST and Fashion MNIST, training generative models sequentially on CIFAR10 is particularly instable, and remains a challenge.

Keywords: Generative Models, Continual Learning

TL;DR: A comparative study of generative models on Continual Learning scenarios.

Code: [ TLESORT/Generative_Continual_Learning](https://github.com/TLESORT/Generative_Continual_Learning)

Data: [MNIST](https://paperswithcode.com/dataset/mnist)

Community Implementations: [ 3 code implementations](https://www.catalyzex.com/paper/arxiv:1812.09111/code)