Data processing: speech signal processing – linguistics – language – Speech signal processing – Recognition
Reexamination Certificate
1999-07-23
2002-05-21
Dorvil, Richemond (Department: 2641)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Recognition
C704S231000
Reexamination Certificate
active
06393396
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention also relates to the art of producing speech suitable for speech recognition from input speech.
The present invention further relates to the art of extracting necessary information from input speech.
The present invention, in addition, relates to the art of processing speech mixed with noise.
2. Description of the Related Art
An interactive system for inputting information based on the recognition of speech uttered by a user and carrying out a process corresponding to the contents of the input information, e.g., outputting a response speech, has been proposed.
Also, in consideration of input speech being spoken in a noise environment, it has been thought to improve the detection rate of user speaking zones or the speech recognition rate by using noise data.
General methods of acquiring noise data are discussed below.
(1) A method of preparing several speech input devices includes placing one of the speech input devices near the user's mouth and the others at places remote from the user's mouth, and acquiring user speech and noise by the separate input devices at the same time.
(2) A method of preparing a noise training mode includes inhibiting the user from speaking, and taking in only noise before the user speaks.
The method (1) is advantageous in acquiring noise data in real time, but the user cannot use a plurality of speech input devices at the same time in many cases. In particular, personal computers that have become increasingly common these days each have only on microphone terminal, and therefore a system employing the method (1) has a difficulty in finding widespread use. The method (2) has the disadvantage that if sudden or peculiar noise is taken in during the noise training mode, noise data different from that produced in the actual environment is stored, and therefore all results of subsequent speech recognition are affected. Also, if noise taken in during the noise training mode differs from actual noise, the results of speech recognition are adversely affected. In addition, a noise training step requires redundant work for the user.
SUMMARY OF THE INVENTION
With the view of solving the above-stated problems in the related art, the present invention intends to realize speech recognition adaptively dealing with surrounding noise even when only one speech input device is employed.
Also, the present invention intends to realize speech recognition adaptively dealing with surrounding noise without a tiresome training work.
REFERENCES:
patent: 5208863 (1993-05-01), Sakurai et al.
patent: 5220629 (1993-06-01), Kosaka et al.
patent: 5276765 (1994-01-01), Freeman et al.
patent: 5369728 (1994-11-01), Kosaka et al.
patent: 5621849 (1997-04-01), Sakurai et al.
patent: 5727072 (1998-03-01), Raman
patent: 5761639 (1998-06-01), Takebayashi et al.
patent: 5787396 (1998-07-01), Komori et al.
patent: 5797116 (1998-08-01), Yamada et al.
patent: 5812975 (1998-09-01), Komoti et al.
patent: 5845047 (1998-12-01), Fukada et al.
patent: 5915234 (1999-06-01), Itoh
patent: 5956679 (1999-09-01), Komori et al.
patent: 5963901 (1999-10-01), Vahatalo et al.
patent: 5970445 (1999-10-01), Yamamoto et al.
patent: 6167375 (2000-12-01), Miseki et al.
Komori Yasuhiro
Kosaka Tetsuo
Nakagawa Ken-ichiro
Dorvil Richemond
Nolan Daniel A.
LandOfFree
Method and apparatus for distinguishing speech from noise does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for distinguishing speech from noise, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for distinguishing speech from noise will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2907032