Meta-learning of Semantic Attention Prototype Correction Based on Bayesian Estimation

Yue Zhang, Baoguo Wei, Lina Zhao, Xinyu Wang, Xu Li, Lixin Li

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

The mean visual prototype is a mainstream meta-learning method for few-shot learning. It is often combined with semantic information to compensate for the lack of training samples. As the number of samples increases, the semantic gain reduces or even becomes noise, leading to a bias in the semantic attention prototype. To address the problem of semantic bias, we propose a prototype correction strategy based on Bayesian estimation. The strategy corrects the original semantic attention prototype defined from the sampling perspective according to the overall information. The probability of the corrected semantic attention prototype is calculated from the probability of the visual prototype and the semantic attention. The semantic attention prototype's posteriori probability center is taken as final prototype. Experiments on Mini-ImageNet and Tiered-ImageNet exhibit our method outperforms existing methods using semantic information by 0.7%~8% in terms of classification accuracy.

源语言英语
主期刊名Proceedings - 2023 11th International Conference on Information Systems and Computing Technology, ISCTech 2023
出版商Institute of Electrical and Electronics Engineers Inc.
311-316
页数6
ISBN(电子版)9798350342406
DOI
出版状态已出版 - 2023
活动11th International Conference on Information Systems and Computing Technology, ISCTech 2023 - Qingdao, 中国
期限: 30 7月 20231 8月 2023

出版系列

姓名Proceedings - 2023 11th International Conference on Information Systems and Computing Technology, ISCTech 2023

会议

会议11th International Conference on Information Systems and Computing Technology, ISCTech 2023
国家/地区中国
Qingdao
时期30/07/231/08/23

指纹

探究 'Meta-learning of Semantic Attention Prototype Correction Based on Bayesian Estimation' 的科研主题。它们共同构成独一无二的指纹。

引用此