Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple Components
Few-shot classification aims at leveraging knowledge learned in a deep learning model, in order to obtain good classification performance on new problems, where only a few labeled samples per class are available. Recent years have seen a fair number of works in the field, each one introducing their...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-06-01
|
Series: | Journal of Imaging |
Subjects: | |
Online Access: | https://www.mdpi.com/2313-433X/8/7/179 |
_version_ | 1811153683920977920 |
---|---|
author | Yassir Bendou Yuqing Hu Raphael Lafargue Giulia Lioi Bastien Pasdeloup Stéphane Pateux Vincent Gripon |
author_facet | Yassir Bendou Yuqing Hu Raphael Lafargue Giulia Lioi Bastien Pasdeloup Stéphane Pateux Vincent Gripon |
author_sort | Yassir Bendou |
collection | DOAJ |
description | Few-shot classification aims at leveraging knowledge learned in a deep learning model, in order to obtain good classification performance on new problems, where only a few labeled samples per class are available. Recent years have seen a fair number of works in the field, each one introducing their own methodology. A frequent problem, though, is the use of suboptimally trained models as a first building block, leading to doubts about whether proposed approaches bring gains if applied to more sophisticated pretrained models. In this work, we propose a simple way to train such models, with the aim of reaching top performance on multiple standardized benchmarks in the field. This methodology offers a new baseline on which to propose (and fairly compare) new techniques or adapt existing ones. |
first_indexed | 2024-03-09T03:19:57Z |
format | Article |
id | doaj.art-f8214e0a0a544c969e7da4f5e7f9b8d0 |
institution | Directory Open Access Journal |
issn | 2313-433X |
language | English |
last_indexed | 2024-03-09T03:19:57Z |
publishDate | 2022-06-01 |
publisher | MDPI AG |
record_format | Article |
series | Journal of Imaging |
spelling | doaj.art-f8214e0a0a544c969e7da4f5e7f9b8d02023-12-03T15:14:20ZengMDPI AGJournal of Imaging2313-433X2022-06-018717910.3390/jimaging8070179Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple ComponentsYassir Bendou0Yuqing Hu1Raphael Lafargue2Giulia Lioi3Bastien Pasdeloup4Stéphane Pateux5Vincent Gripon6IMT Atlantique, Technopole Brest Iroise, 29238 Brest, FranceIMT Atlantique, Technopole Brest Iroise, 29238 Brest, FranceIMT Atlantique, Technopole Brest Iroise, 29238 Brest, FranceIMT Atlantique, Technopole Brest Iroise, 29238 Brest, FranceIMT Atlantique, Technopole Brest Iroise, 29238 Brest, FranceOrange Labs, 35510 Rennes, FranceIMT Atlantique, Technopole Brest Iroise, 29238 Brest, FranceFew-shot classification aims at leveraging knowledge learned in a deep learning model, in order to obtain good classification performance on new problems, where only a few labeled samples per class are available. Recent years have seen a fair number of works in the field, each one introducing their own methodology. A frequent problem, though, is the use of suboptimally trained models as a first building block, leading to doubts about whether proposed approaches bring gains if applied to more sophisticated pretrained models. In this work, we propose a simple way to train such models, with the aim of reaching top performance on multiple standardized benchmarks in the field. This methodology offers a new baseline on which to propose (and fairly compare) new techniques or adapt existing ones.https://www.mdpi.com/2313-433X/8/7/179few-shot learningclassificationdeep learningaugmentationsself-supervisionensembling |
spellingShingle | Yassir Bendou Yuqing Hu Raphael Lafargue Giulia Lioi Bastien Pasdeloup Stéphane Pateux Vincent Gripon Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple Components Journal of Imaging few-shot learning classification deep learning augmentations self-supervision ensembling |
title | Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple Components |
title_full | Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple Components |
title_fullStr | Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple Components |
title_full_unstemmed | Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple Components |
title_short | Easy—Ensemble Augmented-Shot-Y-Shaped Learning: State-of-the-Art Few-Shot Classification with Simple Components |
title_sort | easy ensemble augmented shot y shaped learning state of the art few shot classification with simple components |
topic | few-shot learning classification deep learning augmentations self-supervision ensembling |
url | https://www.mdpi.com/2313-433X/8/7/179 |
work_keys_str_mv | AT yassirbendou easyensembleaugmentedshotyshapedlearningstateoftheartfewshotclassificationwithsimplecomponents AT yuqinghu easyensembleaugmentedshotyshapedlearningstateoftheartfewshotclassificationwithsimplecomponents AT raphaellafargue easyensembleaugmentedshotyshapedlearningstateoftheartfewshotclassificationwithsimplecomponents AT giulialioi easyensembleaugmentedshotyshapedlearningstateoftheartfewshotclassificationwithsimplecomponents AT bastienpasdeloup easyensembleaugmentedshotyshapedlearningstateoftheartfewshotclassificationwithsimplecomponents AT stephanepateux easyensembleaugmentedshotyshapedlearningstateoftheartfewshotclassificationwithsimplecomponents AT vincentgripon easyensembleaugmentedshotyshapedlearningstateoftheartfewshotclassificationwithsimplecomponents |