Speech signal processing apparatus for cutting out a speech signal from a noisy speech signal

• US 5,617,505 A
• Filed: 05/31/1995
• Issued: 04/01/1997
• Est. Priority Date: 05/28/1990
• Status: Expired due to Fees

• Current Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Original Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Inventors: Kane, Joji, Nohara, Akira
• Primary Examiner(s): Tung, Kee M.
• Application Number:
  US08/455,252
• Time in Patent Office: 671 Days
• Field of Search: 395/2, 395/2.18, 395/2.55
• US Class Current: 704/210
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 2021/02087 the noise being separate speech, e.g. cocktail party
• CPC Class Codes: G10L 21/0208 Noise filtering
• CPC Class Codes: G10L 21/0272 Voice signal separating
• CPC Class Codes: G10L 25/12 the extracted parameters being prediction coefficients
• CPC Class Codes: G10L 25/24 the extracted parameters being the cepstrum
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10

Abstract

A band division process including a Fourier transformation is performed for an inputted speech signal, thereby outputting spectrum signals of plural channels. A cepstrum analysis process is performed for the spectrum signals, and a peak of the obtained cepstrum is detected in response to the cepstrum analysis result. Thereafter, a speech signal interval the inputted noisy speech signal is detected in response to the detected peak, and a noise is predicted in the speech signal in response to the detected speech signal interval. Then, the predicted noise is canceled in the spectrum signals thereby outputting noise-suppressed spectrum signals. Finally, the noise-suppressed spectrum signals are combined and are inverse Fourier-transformed, thereby outputting a noise-suppressed speech signal.

Claims

1. A signal processing apparatus comprising:

band division means for performing a band division process including a Fourier transformation for an inputted speech signal and outputting spectrum signals of plural channels;

cepstrum analysis means for performing a cepstrum analysis process for the spectrum signals of plural channels outputted from said band division means and outputting a quefrency value which represents a pitch detected through said cepstrum analysis;

formant analysis means for performing a formant analysis process for the cepstrum analysis result outputted from said cepstrum analysis means and outputting a quefrency value and a cepstrum level which represent a formant detected through said cepstrum analysis;

speech detection means for detecting a feature of a speech of the inputted speech signal based on a combination of the quefrency value which represents the detected pitch outputted from said cepstrum analysis means, and the quefrency value and the cepstrum level which represent the detected formant outputted from said formant analysis means;

speech judgment control means for storing predetermined features of speeches of plural speakers and sequentially outputting the stored features, said predetermined features denoting a predetermined pitch quefrency value and a predetermined formant quefrency value and cepstrum level for each of said plural speakers;

speech judgment means for detecting which speaker among said plural speakers the feature detected by said speech detection means corresponds to by comparing the feature detected by said speech detection means with the stored features outputted from said speech judgment control means and outputting a detection result;

switch control means for outputting a switch control signal according to the detection result outputted from said speech judgment means; and

switch means for outputting speech signals discriminating them by respective speakers according to the switch control signal outputted from said switch control means in response to the outputted speech signal.