EP0474496A2
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Speech recognition apparatus
Assignee: Toshiba Corp
•
Filing: 1991-09-05
•
Granted: —
•
Expires: 2011-09-05
Abstract
Abstract A speech recognition apparatus includes a low-order parameter analysis section (2) for deriving low-order parameter time series of an input speech, a start and end point detection section (3) for detecting the start and end points of the input speech by use of the parameter time series derived by the low-order parameter analysis section, a high-order parameter analysis section (5) for deriving high order parameters at preset sampling points from the input speech of a range corresponding to the start and end points detected by the start and end point detection section, a pattern matching section (7) for matching feature parameters corresponding to the parameters derived by the high-order parameter analysis section with standard parameters previously registered and deriving similarities between the matched parameters, and a recognition result outputting section (8) for outputting the recognition result for the input speech according to the similarities derived by the pattern matching section.
Inventors
- 👤 Yasuki c/o Intellectual Property Div. Yamashita
- 👤 Yoichi C/O Intellectual Property Div Takebayashi
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-10-29 05:07
CPC Classifications
G; G10; G10L; G10L25/00; G10L25/78; G10L25/87; G10L15/00; G10L15/02
Claims Analysis
Total Claims: 13
Claim 1
Independent Claim
125 words
A speech recognition apparatus comprising: low-order parameter analyzation means (2) for deriving low-order parameter time series of an input speech; start and end point detection means (3) for detecting the start and end points of the input speech by use of the parameter time series derived by said low-order parameter analyzation means; high-order parameter analyzation means (5) for deriving high order parameters at preset sampling points from a range of the input speech corresponding to the start and end points detected by said start and end point detection means; and pattern matching means (7) for matching feature parameters corresponding to the parameters derived by said high-order parameter analyzation means with standard parameters previously registered to recognize the input speech according to the result of collation.
Claim 2
Dependent Claim
40 words
A speech recognition apparatus according to claim 1, characterized in that said low-order parameter analyzation means (2) outputs the low order parameter including the result of analyzation containing a small number of arithmetic operations and obtained from the input speech.
Claim 3
Dependent Claim
22 words
A speech recognition apparatus according to claim 2, characterized in that the result of analyzation includes speech power and zero cross number.
Claim 4
Dependent Claim
40 words
A speech recognition apparatus according to claim 1, characterized in that said high-order parameter analyzation means (5) outputs the high order parameter including the result of analyzation containing a large number of arithmetic operations and obtained from the input speech.
Claim 5
Dependent Claim
24 words
A speech recognition apparatus according to claim 4, characterized in that the result of analyzation contains DFT spectrum, filter bank analyzation or LPC parameter.
Claim 6
Dependent Claim
66 words
A speech recognition apparatus according to claim 1, characterized in that said low-order parameter analyzation means (2) derives low order parameters from the input speech signal in a preset range of analyzation window length for each preset analyzation frame period shorter than the analyzation window length and outputs the low order parameter corresponding to the analyzation frame to said start and end point detection means (3).
Claim 7
Dependent Claim
45 words
A speech recognition apparatus according to claim 6, characterized in that said low-order parameter analyzation means (3) derives low order parameters from the input speech signal in a range of the analyzation window length of 24 milliseconds for each analyzation frame period of 8 milliseconds.
Claim 8
Dependent Claim
85 words
A speech recognition apparatus according to claim 6, characterized in that said start and end point detection means (3) includes means for setting a threshold value for the speech energy of the low order parameters and determines the time at which the speech energy of the analyzation frame has exceeded the threshold value as a starting point of a word and the time at which the speech energy of the analyzation frame becomes smaller than the threshold value as an end point of the word.
Claim 9
Dependent Claim
85 words
A speech recognition apparatus according to claim 1, characterized by further comprising a sample frame determination section (4) for determining the frame for analyzing the high order parameters according to the start and end points of said start and end point detection means (3) and in which said high-order parameter analyzation means (5) derives high order parameters including filter analyzation output, DFT spectrum, cepstrum or LPC parameter from the speech signal in a range of the sample frame determined by said sample frame determination section.
Claim 10
Dependent Claim
39 words
A speech recognition apparatus according to claim 1, characterized by further comprising means (6) for calculating word feature parameters from the high order parameters and outputting word feature parameter to said pattern matching means (7) as the feature pattern.
Claim 11
Dependent Claim
176 words
A speech recognition apparatus comprising: low-order parameter analyzation means (2a) for deriving low-order parameter time series from an input speech and outputting speech energy; start and end point candidate detection means (3a) for detecting the candidates of the start and end points of the input speech according to the relation between the level of the speech energy output from said low-order parameter analyzation means and a threshold value; high-order parameter analyzation means (5a) for deriving high order parameters at preset sample points from a range of the input speech corresponding to the candidates of the start and end points detected by said start and end point candidate detection means; start and end point decision means (3b) for checking the relation between the speech energy level and the threshold value and deciding the candidates of the start and end points as the start and end points; and identifying means (7) for identifying the input speech based on the parameters derived by said high-order parameter analyzation means in response to the decision of the start and end points.
Claim 12
Dependent Claim
48 words
A speech recognition apparatus according to claim 11, characterized in that said start and end point decision means (3b) decides the end point candidate as the end point when detecting that the speech energy level is kept larger than the threshold value for a preset period of time.
Claim 13
Dependent Claim
69 words
A speech recognition apparatus according to claim 11, characterized in that said start and end point decision means (3b) includes means for interrupting the analyzation effected by said high-order parameter analyzation means when detecting that a period in which the speech energy level is kept larger than the threshold value is shorter than a preset period of time and operating said start and end point candidate detection means again.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.