氦语音增强算法的研究

A study on the algorithm of helium speech enhancement

当潜水员深海作业时,由于生理方面的原因,需要以氦氧混合气体为呼吸介质替代空气,但这时会出现所谓的“氦语音”现象,语音产生很大畸变,清晰度极低。本文给出了氦语音的基本变化规律,介绍了两种分别基于短时傅立叶变换和线性预测模型的增强算法。我们的实验结果表明,两种算法均能显著校正氦语音,提高清晰度。

When divers work at deep sea, for physiological reasons, they must breathe helium-oxygen mixed atmosphere in place of air. Unfortunately, the resulting helium speech is distorted very much, and nearly unintelligible. This paper describes the essential law of helium speech transformation, presents two enhancement algorithms based respectively on short-term fourier transform and linear prediction. Our experimental results show that both algorithms can correct helium speech and the intelligibility is enhanced greatly.