US6453293B1
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method and device for recognizing at least one keyword in spoken speech using a computer
Assignee: Siemens AG
•
Filing: 2000-11-15
•
Granted: 2002-09-17
•
Expires: 2020-11-15
Abstract
Abstract In order to recognize a keyword in spoken speech, the keyword is subdivided, just like a test pattern to be recognized, into segments. The individual segments of the keyword and of the test pattern are mapped onto one another and a comparison measurement is made and is accumulated over the segments. The keyword is repeatedly stored in a training phase, a plurality of reference features respectively being determined for each segment of the keyword are stored. During recognition, a reference feature wich best fits the relevant segment of the test pattern is assigned in a segment-wise fashion.
Inventors
- 👤 Bernhard Kämmerer
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 22:48
CPC Classifications
G; G10; G10L; G10L15/00; G10L15/04; G10L15/06; G10L15/063; G10L2015/0631; G10L15/08; G10L2015/088
Claims Analysis
Total Claims: 11
Claim 1
Independent Claim
128 words
A speech recognition method, which comprises: using a computer to perform the steps of: storing a keyword in a plurality of representations; subdividing the keyword into key segments; assigning each of the key segments a set of reference features of different representations; subdividing a test pattern derived for spoken speech into test segments; assigning each of the test segments of the test pattern a reference feature from the set of reference features of different representations being most similar to a respective test segment; and recognizing the test pattern as the keyword, if a measure of similarity is determined to be below a prescribed value of an accumulated segment-wise comparison of the reference feature to the respective test segment for each of the test segments of the test pattern.
Claim 2
Dependent Claim
128 words
A speech recognition method, which comprises: using a computer to perform the steps of: storing a keyword in a plurality of representations; subdividing the keyword into key segments; assigning each of the key segments a set of reference features of different representations; subdividing a test pattern derived for spoken speech into test segments; assigning each of the test segments of the test pattern a reference feature from the set of reference features of different representations being most similar to a respective test segment; and recognizing the test pattern as the keyword, if a measure of similarity is determined to be below a prescribed value of an accumulated segment-wise comparison of the reference feature to the respective test segment for each of the test segments of the test pattern.
Claim 3
Dependent Claim
18 words
The method according to claim 1 , which comprises using an independent sound unit as the test pattern.
Claim 4
Dependent Claim
24 words
The method according to claim 1 , which comprises setting a number of the keys segments equal to a number of the test segments.
Claim 5
Dependent Claim
29 words
The method according to claim 1 , which comprises comparing the test pattern with a plurality of keywords, and outputting the keyword being most similar to the test pattern.
Claim 6
Dependent Claim
55 words
The method according to claim 1 , which comprises using feature vectors for storing the spoken speech representing the keyword and the test pattern, respectively, in which case at prescribed sampling instants the keyword and the test pattern are digitized, respectively, and one of the feature vectors is stored with data characterizing the spoken speech.
Claim 7
Dependent Claim
42 words
The method according to claim 5 , which comprises storing for each segment of the key segments and of the test segments, a feature vector which is averaged over all the feature vectors of the segment, the feature vector characterizes the segment.
Claim 8
Dependent Claim
24 words
The method according to claim 1 , which comprises determining the prescribed value such that a best fit for the keyword is always recognized.
Claim 9
Dependent Claim
17 words
The method as claimed in claim 1 , which comprises using a word as the test pattern.
Claim 10
Dependent Claim
113 words
A speech recognition device, comprising: a processor unit programmed to: store a keyword in a plurality of representations; subdivide a said keyword into key segments; assign each of said key segments a set of reference features of different representations; subdivide a test pattern of spoken speech into test segments; assign each of said test segments of said test pattern a reference feature from said set of reference features of different representations being most similar to a respective test segment; and recognize said test pattern as said keyword, if a measure of similarity is determined to be below a prescribed value of an accumulated segment-wise comparison of said reference feature to said test segment.
Claim 11
Dependent Claim
113 words
A speech recognition device, comprising: a processor unit programmed to: store a keyword in a plurality of representations; subdivide a said keyword into key segments; assign each of said key segments a set of reference features of different representations; subdivide a test pattern of spoken speech into test segments; assign each of said test segments of said test pattern a reference feature from said set of reference features of different representations being most similar to a respective test segment; and recognize said test pattern as said keyword, if a measure of similarity is determined to be below a prescribed value of an accumulated segment-wise comparison of said reference feature to said test segment.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2005-09-16 | 2006-03-18 | 2006-09-17 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2009-09-16 | 2010-03-18 | 2010-09-17 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2013-09-16 | 2014-03-18 | 2014-09-17 | $7700.00 | Window Closed / Lapsed |