IPNET:Influential Prototypical Networks for Few Shot Learning

TL;DR

This paper introduces IPNET, an improved prototypical network for few-shot learning that weights support samples by their influence on class distribution, enhancing classification accuracy.

Contribution

It proposes a novel influence-weighted approach to prototypical networks using MMD to better represent class distributions in few-shot learning.

Findings

Improved classification performance over standard PN.

Effective weighting of support samples based on influence.

Enhanced class representation accuracy.

Abstract

Prototypical network (PN) is a simple yet effective few shot learning strategy. It is a metric-based meta-learning technique where classification is performed by computing Euclidean distances to prototypical representations of each class. Conventional PN attributes equal importance to all samples and generates prototypes by simply averaging the support sample embeddings belonging to each class. In this work, we propose a novel version of PN that attributes weights to support samples corresponding to their influence on the support sample distribution. Influence weights of samples are calculated based on maximum mean discrepancy (MMD) between the mean embeddings of sample distributions including and excluding the sample. Further, the influence factor of a sample is measured using MMD based on the shift in the distribution in the absence of that sample.