Word Embedding Distribution Propagation Graph Network for Few-Shot Learning
Few-shot learning (FSL) is of great significance to the field of machine learning. The ability to learn and generalize using a small number of samples is an obvious distinction between artificial intelligence and humans. In the FSL domain, most graph neural networks (GNNs) focus on transferring labe...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/7/2648 |
| _version_ | 1827622403108241408 |
|---|---|
| author | Chaoran Zhu Ling Wang Cheng Han |
| author_facet | Chaoran Zhu Ling Wang Cheng Han |
| author_sort | Chaoran Zhu |
| collection | DOAJ |
| description | Few-shot learning (FSL) is of great significance to the field of machine learning. The ability to learn and generalize using a small number of samples is an obvious distinction between artificial intelligence and humans. In the FSL domain, most graph neural networks (GNNs) focus on transferring labeled sample information to an unlabeled query sample, ignoring the important role of semantic information during the classification process. Our proposed method embeds semantic information of classes into a GNN, creating a word embedding distribution propagation graph network (WPGN) for FSL. We merge the attention mechanism with our backbone network, use the Mahalanobis distance to calculate the similarity of classes, select the Funnel ReLU (FReLU) function as the activation function of the Transform layer, and update the point graph and word embedding distribution graph. In extensive experiments on FSL benchmarks, compared with the baseline model, the accuracy of the WPGN on the 5-way-1/2/5 shot tasks increased by 9.03, 4.56, and 4.15%, respectively. |
| first_indexed | 2024-03-09T11:25:02Z |
| format | Article |
| id | doaj.art-1e083e28eba74436bbb5d1ce592579a4 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T11:25:02Z |
| publishDate | 2022-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-1e083e28eba74436bbb5d1ce592579a42023-12-01T00:02:44ZengMDPI AGSensors1424-82202022-03-01227264810.3390/s22072648Word Embedding Distribution Propagation Graph Network for Few-Shot LearningChaoran Zhu0Ling Wang1Cheng Han2College of Computer Science and Technology, Changchun University of Science and Technology, Changchun 130022, ChinaCollege of Computer Science and Technology, Changchun University of Science and Technology, Changchun 130022, ChinaCollege of Computer Science and Technology, Changchun University of Science and Technology, Changchun 130022, ChinaFew-shot learning (FSL) is of great significance to the field of machine learning. The ability to learn and generalize using a small number of samples is an obvious distinction between artificial intelligence and humans. In the FSL domain, most graph neural networks (GNNs) focus on transferring labeled sample information to an unlabeled query sample, ignoring the important role of semantic information during the classification process. Our proposed method embeds semantic information of classes into a GNN, creating a word embedding distribution propagation graph network (WPGN) for FSL. We merge the attention mechanism with our backbone network, use the Mahalanobis distance to calculate the similarity of classes, select the Funnel ReLU (FReLU) function as the activation function of the Transform layer, and update the point graph and word embedding distribution graph. In extensive experiments on FSL benchmarks, compared with the baseline model, the accuracy of the WPGN on the 5-way-1/2/5 shot tasks increased by 9.03, 4.56, and 4.15%, respectively.https://www.mdpi.com/1424-8220/22/7/2648few-shot learninggraph neural networksemantic informationattention mechanismMahalanobis distanceFReLU |
| spellingShingle | Chaoran Zhu Ling Wang Cheng Han Word Embedding Distribution Propagation Graph Network for Few-Shot Learning Sensors few-shot learning graph neural network semantic information attention mechanism Mahalanobis distance FReLU |
| title | Word Embedding Distribution Propagation Graph Network for Few-Shot Learning |
| title_full | Word Embedding Distribution Propagation Graph Network for Few-Shot Learning |
| title_fullStr | Word Embedding Distribution Propagation Graph Network for Few-Shot Learning |
| title_full_unstemmed | Word Embedding Distribution Propagation Graph Network for Few-Shot Learning |
| title_short | Word Embedding Distribution Propagation Graph Network for Few-Shot Learning |
| title_sort | word embedding distribution propagation graph network for few shot learning |
| topic | few-shot learning graph neural network semantic information attention mechanism Mahalanobis distance FReLU |
| url | https://www.mdpi.com/1424-8220/22/7/2648 |
| work_keys_str_mv | AT chaoranzhu wordembeddingdistributionpropagationgraphnetworkforfewshotlearning AT lingwang wordembeddingdistributionpropagationgraphnetworkforfewshotlearning AT chenghan wordembeddingdistributionpropagationgraphnetworkforfewshotlearning |