语音情感识别在联合收割机疲劳驾驶预警中的研究——基于汉语语言音韵
Speech Emotion Recognition in Driving Fatigue Warning of Combine Harvester Based on Chinese Language Phonology
胡士强 1罗杨2
作者信息
- 1. 红河学院,云南 红河 661100
- 2. 宜宾学院,四川 宜宾 644000
- 折叠
摘要
通过汉语语言音韵的语音情感识别实现联合收割机疲劳驾驶检测具有非侵入性、高效准确、实时性强、可普及性强和应用场景广泛等优点.因此,为了提高疲劳驾驶准确识别率,将梅尔频率倒谱系数和汉语语言音韵融合在一起,利用高斯混合模型和极限学习机作为分类器进行语音疲劳检测.检测结果表明:该方法疲劳识别率很高,平均正确识别率达到了 93.17%,具有较高的实用价值和推广价值.
Abstract
It has the advantages of non-invasive,efficient and accurate,strong real-time,strong popularity,and wide application scenarios for the realization of combine fatigue driving detection through speech emotion recognition of Chinese language phonology.In order to improve the accurate recognition rate of fatigue driving,it combined Mel Frequency Cep-strum Coefficients and Chinese language phonology,and used Gaussian mixture model and limit learning machine as clas-sifiers for speech fatigue detection.The experimental results showed that the fatigue identification rate of this method was very high,with an average correct identification rate of 93.17%,which had high practical value and promotion value.
关键词
汉语语言音韵/情感识别/联合收割机/疲劳驾驶Key words
Chinese language phonology/emotional recognition/combine harvester/fatigue driving引用本文复制引用
出版年
2025
国家科技期刊平台
NSTL
万方数据