KGTN-ens: few-shot image classification with knowledge graph ensembles

Dominik Filipiak

doi:10.1007/s10489-023-05129-8

KGTN-ens: few-shot image classification with knowledge graph ensembles

2024-01-25

1

OA

AI

DOI：10.1007/s10489-023-05129-8

Original

Original request for help

Save

Abstract

En 中文

We propose KGTN-ens, a framework extending the recent Knowledge GraphTransferNetwork (KGTN) in order to incorporate multiple knowledge graph embeddings at a small cost. There are many real-world scenarios in which the amount of data is severely limited (e.g. health industry, rare anomalies). Prior knowledge can be used to tackle this task. In KGTN, one can use a single knowledge source at once. The purpose of this study is to investigate the possibility of combining multiple knowledge sources. We evaluate it with different embeddings in a few-shot image classification task. Our model is partially trained on k. {1, 2, 5, 10} samples. We also construct a new knowledge source - Wikidata embeddings - and evaluate it with KGTN and KGTN-ens. With ResNet50, our approach outperforms KGTN in terms of the top-5 accuracy on the ImageNet-FS dataset for the majority of tested settings. For k. {1, 2, 5, 10} respectively, we obtained +0.63/+0.58/+0.43/+0.26 pp. (novel classes) and +0.26/+0.25/+0.32/-0.04 pp. (all classes).

Keywords:

Few-shot Image Classification

Knowledge Graph Enabled AI

Ensemble Learning

KGTN-ens: few-shot image classification with knowledge graph ensembles

Abstract

Journal

Organization