论文标题
kgtn-ens：知识图合奏的很少的图像分类
KGTN-ens: Few-Shot Image Classification with Knowledge Graph Ensembles
论文作者
论文摘要
我们提出了KGTN-ens，这是一个扩展了最新知识图传输网络（KGTN）的框架，以便以较小的成本结合多个知识图嵌入。我们在几个弹奏图像分类任务中以不同的嵌入组合进行评估。我们还构建了一个新的知识源-Wikidata嵌入 - 并使用KGTN和KGTN -ENS进行评估。对于大多数测试设置，我们的方法在Imagenet-FS数据集上的前5个精度上优于KGTN。
We propose KGTN-ens, a framework extending the recent Knowledge Graph Transfer Network (KGTN) in order to incorporate multiple knowledge graph embeddings at a small cost. We evaluate it with different combinations of embeddings in a few-shot image classification task. We also construct a new knowledge source - Wikidata embeddings - and evaluate it with KGTN and KGTN-ens. Our approach outperforms KGTN in terms of the top-5 accuracy on the ImageNet-FS dataset for the majority of tested settings.
