Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Graph-based Interpolation of Feature Vectors for Accurate Few-Shot Classification

About

In few-shot classification, the aim is to learn models able to discriminate classes using only a small number of labeled examples. In this context, works have proposed to introduce Graph Neural Networks (GNNs) aiming at exploiting the information contained in other samples treated concurrently, what is commonly referred to as the transductive setting in the literature. These GNNs are trained all together with a backbone feature extractor. In this paper, we propose a new method that relies on graphs only to interpolate feature vectors instead, resulting in a transductive learning setting with no additional parameters to train. Our proposed method thus exploits two levels of information: a) transfer features obtained on generic datasets, b) transductive information obtained from other samples to be classified. Using standard few-shot vision classification datasets, we demonstrate its ability to bring significant gains compared to other works.

Yuqing Hu, Vincent Gripon, St\'ephane Pateux• 2020

Related benchmarks

TaskDatasetResultRank
Few-shot classificationmini-ImageNet → CUB (test)
Accuracy (5-shot)73.46
75
Few-shot Image ClassificationminiImageNet (novel classes)--
20
Few-shot Image ClassificationCUB (novel)
1-Shot Acc88.35
14
Few-shot Image ClassificationCIFAR-FS (novel)
1-shot Acc83.9
7
Showing 4 of 4 rows

Other info

Code

Follow for update