Feature Extractor Stacking for Cross-domain Few-shot Learning

TL;DR

This paper introduces feature extractor stacking (FES), a novel method for cross-domain few-shot learning that combines heterogeneous pretrained extractors without needing to re-compute a universal model, achieving state-of-the-art results.

Contribution

The paper proposes FES, a flexible stacking approach for CDFSL that handles heterogeneous extractors and updates efficiently, unlike previous universal model methods.

Findings

FES achieves state-of-the-art performance on Meta-Dataset.

FES effectively combines diverse pretrained extractors.

FES outperforms existing CDFSL methods in accuracy.

Abstract

Cross-domain few-shot learning (CDFSL) addresses learning problems where knowledge needs to be transferred from one or more source domains into an instance-scarce target domain with an explicitly different distribution. Recently published CDFSL methods generally construct a universal model that combines knowledge of multiple source domains into one feature extractor. This enables efficient inference but necessitates re-computation of the extractor whenever a new source domain is added. Some of these methods are also incompatible with heterogeneous source domain extractor architectures. We propose feature extractor stacking (FES), a new CDFSL method for combining information from a collection of extractors, that can utilise heterogeneous pretrained extractors out of the box and does not maintain a universal model that needs to be re-computed when its extractor collection is updated. We present the basic FES algorithm, which is inspired by the classic stacked generalisation approach, and also introduce two variants: convolutional FES (ConFES) and regularised FES (ReFES). Given a target-domain task, these algorithms fine-tune each extractor independently, use cross-validation to extract training data for stacked generalisation from the support set, and learn a simple linear stacking classifier from this data. We evaluate our FES methods on the well-known Meta-Dataset benchmark, targeting image classification with convolutional neural networks, and show that they can achieve state-of-the-art performance.