Skip to Main content Skip to Navigation

Selecting Relevant Features from a Universal Representation for Few-shot Classification

Nikita Dvornik 1 Cordelia Schmid 1 Julien Mairal 1
1 Thoth - Apprentissage de modèles à partir de données massives
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann
Abstract : Popular approaches for few-shot classification consist of first learning a generic data representation based on a large annotated dataset, before adapting the representation to new classes given only a few labeled samples. In this work, we propose a new strategy based on feature selection, which is both simpler and more effective than previous feature adaptation approaches. First, we obtain a universal representation by training a set of semantically different feature extractors. Then, given a few-shot learning task, we use our universal feature bank to automatically select the most relevant representations. We show that a simple non-parametric classifier built on top of such features produces high accuracy and generalizes to domains never seen during training, which leads to state-of-the-art results on MetaDataset and improved accuracy on mini-ImageNet.
Document type :
Preprints, Working Papers, ...
Complete list of metadatas

Cited literature [57 references]  Display  Hide  Download
Contributor : Nikita Dvornik <>
Submitted on : Sunday, March 22, 2020 - 2:53:43 PM
Last modification on : Thursday, March 26, 2020 - 8:50:00 PM


Files produced by the author(s)


  • HAL Id : hal-02513241, version 2
  • ARXIV : 2003.09338



Nikita Dvornik, Cordelia Schmid, Julien Mairal. Selecting Relevant Features from a Universal Representation for Few-shot Classification. 2020. ⟨hal-02513241v2⟩



Record views


Files downloads