# Class Incremental Continual Learning with Self-Organizing Maps and Variational Autoencoders Using Synthetic Replay

## TL;DR

This paper presents a scalable, memory-efficient continual learning framework using self-organizing maps and variational autoencoders with synthetic replay, achieving competitive results on standard benchmarks without storing raw data.

## Contribution

It introduces a novel generative continual learning approach combining SOMs and VAEs that operates in latent space, eliminating the need for raw data storage and improving performance.

## Key findings

- Outperforms memory-free methods on CIFAR-10 and CIFAR-100.
- Achieves nearly 10% and 7% improvements over state-of-the-art in class-incremental learning.
- Provides scalable, task-label-free continual learning with visualization capabilities.

## Abstract

This work introduces a novel generative continual learning framework based on self-organizing maps (SOMs) and variational autoencoders (VAEs) to enable memory-efficient replay, eliminating the need to store raw data samples or task labels. For high-dimensional input spaces, such as of CIFAR-10 and CIFAR-100, we design a scheme where the SOM operates over the latent space learned by a VAE, whereas, for lower-dimensional inputs, such as those found in MNIST and FashionMNIST, the SOM operates in a standalone fashion. Our method stores a running mean, variance, and covariance for each SOM unit, from which synthetic samples are then generated during future learning iterations. For the VAE-based method, generated samples are then fed through the decoder to then be used in subsequent replay. Experimental results on standard class-incremental benchmarks show that our approach performs competitively with state-of-the-art memory-based methods and outperforms memory-free methods, notably improving over best state-of-the-art single class incremental performance on CIFAR-10 and CIFAR-100 by nearly $10$\% and $7$\%, respectively. Our methodology further facilitates easy visualization of the learning process and can also be utilized as a generative model post-training. Results show our method's capability as a scalable, task-label-free, and memory-efficient solution for continual learning.