Few-Shot Classification with Feature Map Reconstruction Networks

TL;DR

This paper proposes a novel few-shot classification method that reconstructs query features from support features in latent space, improving accuracy and efficiency without adding complex modules.

Contribution

It introduces a feature map reconstruction approach that directly regresses from support to query features in closed form, enhancing performance and computational efficiency.

Findings

Achieves significant accuracy improvements on four fine-grained benchmarks.

Outperforms previous methods in efficiency and accuracy.

Competitive results on mini-ImageNet and tiered-ImageNet benchmarks.

Abstract

In this paper we reformulate few-shot classification as a reconstruction problem in latent space. The ability of the network to reconstruct a query feature map from support features of a given class predicts membership of the query in that class. We introduce a novel mechanism for few-shot classification by regressing directly from support features to query features in closed form, without introducing any new modules or large-scale learnable parameters. The resulting Feature Map Reconstruction Networks are both more performant and computationally efficient than previous approaches. We demonstrate consistent and substantial accuracy gains on four fine-grained benchmarks with varying neural architectures. Our model is also competitive on the non-fine-grained mini-ImageNet and tiered-ImageNet benchmarks with minimal bells and whistles.