电子学报2023，Vol.51Issue(10)：2915-2924,10.DOI:10.12263/DZXB.20220289
基于特征空间轨迹信息的语音关键词检测方法
Spoken Term Detection Based on Feature Space Trajectory Information
摘要
Abstract
The current technique of spoken term detection is dominated by deep learning,which requires large anno-tated data for training,and is difficult to be applied in limited-data scenarios.In this paper,a feature trajectory based meth-od of spoken term detection is proposed for limited-data scenarios.The method originated from the fact that a word is a structured organization of small units such as syllable or phoneme and any language unit has steady statistical audio feature,based on the principle of physical location,feature distribution,temporal information of keywords,and local distinguishing information are constructed with speech examples.Spoken keywords are searched with the feature trajectory information of the detected speech segment in hierarchical decision strategy.The method works on a audio feature space defined by a iden-tifier set trained with a large unlabeled speech dataset.Several experimental results show that the proposed method is evi-dently superior to HMM and CRNN when the training samples is less than 100.For example,when 10 samples are used for training,FRR and FAR of the propose method are absolutely decreased by 20.5％and 8.7 FP/hour respectively compared with HMM-based system.On the other hand,the proposed method achieved the comparable performance v.s.CRNN-based system when the training samples is more than 300.关键词
语音关键词检测/音频特征空间/特征空间轨迹信息/低资源Key words
spoken term detection/audio feature space/feature space trajectory information/limited-data source分类
信息技术与安全科学引用本文复制引用
田颖慧,贺前华,郑若伟,危卓,李艳雄..基于特征空间轨迹信息的语音关键词检测方法[J].电子学报,2023,51(10):2915-2924,10.基金项目
广东省自然科学基金(No.2022A1515011687) （No.2022A1515011687）
国家自然科学基金(No.61571192)Guangdong Natural Science Foundation(No.2022A1515011687) （No.61571192）
National Nature Science Foundation of China(No.61571192) （No.61571192）
