Image Classification with Small Datasets: Overview and Benchmark

TL;DR

This paper provides a comprehensive overview and benchmark for image classification methods on small datasets, revealing that hyper-parameter tuning significantly impacts performance and only one specialized method outperforms the baseline.

Contribution

It introduces a systematic overview and a new benchmark for small dataset image classification, enabling objective comparison of methods.

Findings

Thorough hyper-parameter tuning yields a highly competitive baseline.

Performance growth over recent years has been limited.

Only one specialized method from 2019 outperforms the baseline.

Abstract

Image classification with small datasets has been an active research area in the recent past. However, as research in this scope is still in its infancy, two key ingredients are missing for ensuring reliable and truthful progress: a systematic and extensive overview of the state of the art, and a common benchmark to allow for objective comparisons between published methods. This article addresses both issues. First, we systematically organize and connect past studies to consolidate a community that is currently fragmented and scattered. Second, we propose a common benchmark that allows for an objective comparison of approaches. It consists of five datasets spanning various domains (e.g., natural images, medical imagery, satellite data) and data types (RGB, grayscale, multispectral). We use this benchmark to re-evaluate the standard cross-entropy baseline and ten existing methods published between 2017 and 2021 at renowned venues. Surprisingly, we find that thorough hyper-parameter tuning on held-out validation data results in a highly competitive baseline and highlights a stunted growth of performance over the years. Indeed, only a single specialized method dating back to 2019 clearly wins our benchmark and outperforms the baseline classifier.