听觉场景分析(Auditory Scene Analysis,ASA)系统能将一个场景分解为与不同声源对应的语音流。分割是ASA的主要步骤,借助分割可将一个听觉场景分解成多个片断。实现基于上升缘和下降缘分析的语音分割系统需检测上升缘与下降缘,通过匹配...听觉场景分析(Auditory Scene Analysis,ASA)系统能将一个场景分解为与不同声源对应的语音流。分割是ASA的主要步骤,借助分割可将一个听觉场景分解成多个片断。实现基于上升缘和下降缘分析的语音分割系统需检测上升缘与下降缘,通过匹配对应的上升缘与下降缘的波前来生成语音片断,将这些片断重构成语音流。展开更多
This letter proposes a new method for concurrent voiced speech separation. Firstly the Wrapped Discrete Fourier Transform (WDFT) is used to decompose the harmonic spectra of the mixed speeches. Then the individual spe...This letter proposes a new method for concurrent voiced speech separation. Firstly the Wrapped Discrete Fourier Transform (WDFT) is used to decompose the harmonic spectra of the mixed speeches. Then the individual speech is reconstructed by using the sinusoidal speech model. By taking advantage of the non-uniform frequency resolution of WDFT, harmonic spectra parameters can be estimated and separated accurately. Experimental results on mixed vowels separation show that the proposed method can recover the original speeches effectively.展开更多
文摘听觉场景分析(Auditory Scene Analysis,ASA)系统能将一个场景分解为与不同声源对应的语音流。分割是ASA的主要步骤,借助分割可将一个听觉场景分解成多个片断。实现基于上升缘和下降缘分析的语音分割系统需检测上升缘与下降缘,通过匹配对应的上升缘与下降缘的波前来生成语音片断,将这些片断重构成语音流。
基金Supported by the National Natural Science Foundation of China (No.60172048).
文摘This letter proposes a new method for concurrent voiced speech separation. Firstly the Wrapped Discrete Fourier Transform (WDFT) is used to decompose the harmonic spectra of the mixed speeches. Then the individual speech is reconstructed by using the sinusoidal speech model. By taking advantage of the non-uniform frequency resolution of WDFT, harmonic spectra parameters can be estimated and separated accurately. Experimental results on mixed vowels separation show that the proposed method can recover the original speeches effectively.