# Few-shot Image Classification: Just Use a Library of Pre-trained Feature Extractors and a Simple Classifier

Source: [Few-shot Image Classification: Just Use a Library of Pre-trained Feature Extractors and a Simple Classifier](<https://www.sota2.com/research/paper/68fa6a981e209131cea29317>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-27T06:09:35.928Z

Authors: Arkabandhu Chowdhury, Mingchao Jiang, Swarat Chaudhuri, Chris Jermaine

Published: 2021-01-03T05:30:36.000Z

## Abstract

Recent papers have suggested that transfer learning can outperform sophisticated meta-learning methods for few-shot image classification. We take this hypothesis to its logical conclusion, and suggest the use of an ensemble of high-quality, pre-trained feature extractors for few-shot image classification. We show experimentally that a library of pre-trained feature extractors combined with a simple feed-forward network learned with an L2-regularizer can be an excellent option for solving cross-domain few-shot image classification. Our experimental results suggest that this simpler sample-efficient approach far outperforms several well-established meta-learning algorithms on a variety of few-shot tasks.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Image Classification on PlantVillage 1-shot (test)](<https://www.sota2.com/research/sota/image-classification-on-plantvillage-1-shot-test.md>): Image Classification; PlantVillage 1-shot (test); Accuracy: 37.89.

- [Image Classification on PlantVillage 5-shot (test)](<https://www.sota2.com/research/sota/image-classification-on-plantvillage-5-shot-test.md>): Image Classification; PlantVillage 5-shot (test); Accuracy: 62.07.

- [Image Classification on PlantVillage 10-shot (test)](<https://www.sota2.com/research/sota/image-classification-on-plantvillage-10-shot-test.md>): Image Classification; PlantVillage 10-shot (test); Accuracy: 72.33.

- [Image Classification on PlantVillage 15-shot (test)](<https://www.sota2.com/research/sota/image-classification-on-plantvillage-15-shot-test.md>): Image Classification; PlantVillage 15-shot (test); Accuracy: 76.93.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2101.00562>)

[PDF](<https://arxiv.org/pdf/2101.00562.pdf>)