DeepCore: A Comprehensive Library for Coreset Selection in Deep Learning

TL;DR

DeepCore is a comprehensive library for coreset selection in deep learning, providing empirical evaluations of various methods on CIFAR10 and ImageNet, highlighting the strength of random selection as a baseline.

Contribution

The paper introduces DeepCore, a library for coreset selection in deep learning, and offers an extensive empirical study comparing existing methods on standard datasets.

Findings

Random selection remains a strong baseline.

Existing methods have advantages in specific settings.

Coreset selection methods vary in effectiveness depending on the scenario.

Abstract

Coreset selection, which aims to select a subset of the most informative training samples, is a long-standing learning problem that can benefit many downstream tasks such as data-efficient learning, continual learning, neural architecture search, active learning, etc. However, many existing coreset selection methods are not designed for deep learning, which may have high complexity and poor generalization performance. In addition, the recently proposed methods are evaluated on models, datasets, and settings of different complexities. To advance the research of coreset selection in deep learning, we contribute a comprehensive code library, namely DeepCore, and provide an empirical study on popular coreset selection methods on CIFAR10 and ImageNet datasets. Extensive experiments on CIFAR10 and ImageNet datasets verify that, although various methods have advantages in certain experiment settings, random selection is still a strong baseline.