One-shot Learning with Absolute Generalization

TL;DR

This paper introduces the concept of absolute generalization in datasets and proposes a method to build classifiers capable of one-shot learning across such datasets, transforming classification into identity or similarity tasks.

Contribution

It defines dataset conditions for one-shot learning support and presents a novel classifier construction method based on sample concatenation and similarity measurement.

Findings

The proposed method outperforms baselines on one-shot learning datasets.

It effectively transforms classification into identity or similarity problems.

Experiments validate the superiority of the approach on artificial datasets.

Abstract

One-shot learning is proposed to make a pretrained classifier workable on a new dataset based on one labeled samples from each pattern. However, few of researchers consider whether the dataset itself supports one-shot learning. In this paper, we propose a set of definitions to explain what kind of datasets can support one-shot learning and propose the concept "absolute generalization". Based on these definitions, we proposed a method to build an absolutely generalizable classifier. The proposed method concatenates two samples as a new single sample, and converts a classification problem to an identity identification problem or a similarity metric problem. Experiments demonstrate that the proposed method is superior to baseline on one-shot learning datasets and artificial datasets.