Recommendation of Data-Free Class-Incremental Learning Algorithms by Simulating Future Data

Eva Feillet, Adrian Popescu, Céline Hudelot

Abstract: Class-incremental learning deals with data streams composed of batches of classes. Various algorithms have been proposed to address the challenging case where samples from past classes cannot be stored. However, selecting an appropriate algorithm for a user-defined setting is an open problem, as the relative performance of these algorithms depends on the incremental setting. To solve this problem, we introduce an algorithm recommendation method that simulates the future data stream. Given an initial set of classes, our method leverages generative models to simulate future classes from the same visual domain. We evaluate recent algorithms on the simulated stream and recommend the one that performs best in the user-defined incremental setting. We illustrate the effectiveness of our method on three large datasets using six algorithms and six incremental settings. Our method performs close to an oracle that would choose the best algorithm in each setting. This work contributes to facilitating the practical deployment of continual learning.