摘要
情感是语音识别研究中一个不可避免的问题,不同的情感对于语音有着不同的影响,这种影响使得中性语音识别系统在实际应用中的识别效果大打折扣。对于类似的影响通常的解决方法有寻找鲁棒特征,特征归一化以及模型调整训练等。本文通过自适应方法,使用少量情感数据,在中性语音模型的基础上自适应得到新的情感语音模型。实验证明,新模型对于情感语音有着更好的整体识别率。
As known to all, emotion plays a significant role in speech recognition. The model built on neutral speech degrades dramatically while recognizing speech with emotion. How to deal with emotion issue properly is crucial to achieve good performance in recognition. Most widely used approaches include robust feature extraction, speaker normalization and model tuning/retraining. In the study, a novel method is proposed, that is, adaptation technique is adopted to transfbrm a neutral-based model into emotion-specific one with a small amount of emotion speech. It＇s shown experimentally that the new model achieves higher accuracy in overall performance.
出处
《计算机科学》
CSCD
北大核心
2007年第1期163-165,共3页
Computer Science
基金
国家自然科学基金重点项目-情感计算理论与方法研究(编号:60433030
分类代码:F020106)
关键词
语音识别
情感语音识别
情感计算
自适应
Speech recognition, Emotional speech recognition, Affective computing, Adaptation
作者简介
潘玉春 硕士研究生。
