【摘要】 提出了一种基于谱减法的语音增强算法。算法利用人耳的听觉掩蔽效应动态确定每一帧的权值,有选择地增强。实验结果表明,方法有效抑制了背景噪声和音乐噪声并提高了语音质量。更多还原

【Abstract】 An approach to speech-sound enhancement is proposed based on the spectral subtraction.The masking of the auditory system is used to determine the value of each frame dynamically and the enhancement is performed selectively.The experiment result shows that the approach restrains from the background noise and music residual noise and improved the speech-sound quality.更多还原