Patent US5864804A successfully refreshed from Google Patents.
US5864804A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Voice recognition system
Assignee: US Philips Corp
•
Filing: 1996-06-10
•
Granted: 1999-01-26
•
Expires: 2016-06-10
Abstract
Abstract The invention relates to a voice recognition system which is robust to echoes and other background noise. An additional input signal describing a disturbance is evaluated such that during this additional recognition there is maximum suppression of information contained in the input signal. For this purpose, comparison vectors are formed which are continuously adapted to the instantaneous interference. The voice recognition system receives speech signals superimposed by noise signals. Additionally, a first spectral analysis unit produces first spectral values combined to first spectral vectors derived from disturbed speech signals. Estimates of the noise signals are also produced. A second spectral analysis unit produces second spectral values combined to second spectral vectors from the noise signal estimates. Reference vectors formed by reference spectral values to which reference vectors are assigned indices and which reference vectors respond to the spectrums of reference speech signals are stored in memory. Comparison vectors are produced and assigned to the indices from the reference vectors and the second spectral vectors, a continuous adaptation being effected of at least one of the comparison vectors to the second spectral vectors. Indices are then selected by comparing the comparison vectors with the first spectral vectors.
Inventors
- 👤 Hans Kalveram
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 21:19
CPC Classifications
G; G10; G10L; G10L15/00; G10L15/20; G10L21/00; G10L21/02; G10L21/0208; G10L21/0216
Claims Analysis
Total Claims: 15
Claim 1
Independent Claim
165 words
A voice recognition system, comprising: first spectral analysis means for receiving disturbed speech signals (X) composed of speech signals (S) superimposed by noise signals (N), and for producing first spectral vectors (x) comprising first spectral values (xd) from the disturbed speech signals (X); means for producing estimates (N) of the noise signals (N); second spectral analysis means for producing second spectral vectors (n) comprising second spectral values (nd) from the noise signal estimates (N); a memory unit for storing reference vectors (y) comprising reference spectral values (yd), the reference vectors (y) being assigned indices (j), wherein the reference vectors (y) correspond to spectrums of reference speech signals; means for producing comparison vectors (y') assigned to the indices (j) from the reference vectors (y) and the second spectral vectors (n), and for continuously adapting at least one of the comparison vectors (y') to the second spectral vectors (n); and means for selecting indices (j) by comparing the comparison vectors (y') with the first spectral vectors (x).
Claim 2
Dependent Claim
165 words
A voice recognition system, comprising: first spectral analysis means for receiving disturbed speech signals (X) composed of speech signals (S) superimposed by noise signals (N), and for producing first spectral vectors (x) comprising first spectral values (xd) from the disturbed speech signals (X); means for producing estimates (N) of the noise signals (N); second spectral analysis means for producing second spectral vectors (n) comprising second spectral values (nd) from the noise signal estimates (N); a memory unit for storing reference vectors (y) comprising reference spectral values (yd), the reference vectors (y) being assigned indices (j), wherein the reference vectors (y) correspond to spectrums of reference speech signals; means for producing comparison vectors (y') assigned to the indices (j) from the reference vectors (y) and the second spectral vectors (n), and for continuously adapting at least one of the comparison vectors (y') to the second spectral vectors (n); and means for selecting indices (j) by comparing the comparison vectors (y') with the first spectral vectors (x).
Claim 3
Dependent Claim
35 words
The voice recognition system as claimed in claim 1, wherein noise signals (N) are independently captured in the neighborhood of an associated noise signal source to form the estimates (N) of the noise signals (N).
Claim 4
Dependent Claim
29 words
The voice recognition system as claimed in claim 1 wherein echoes of audio signals known a priori are copied to form the estimates (N) of the noise signals (N).
Claim 5
Dependent Claim
52 words
The voice recognition system as claimed in claim 1, wherein an adaptive filter is used for filtering signals (N) of an associated noise signal source to form the estimates (N) of the noise signals (N); and further comprising means for subtracting the estimates (N) thus formed from the disturbed speech signals (X).
Claim 6
Dependent Claim
31 words
The voice recognition system as claimed in claim 1 wherein the comparison vector (y') corresponding to a speech signal pause (j=0) is continuously replaced by the present second spectral vector (n).
Claim 7
Dependent Claim
24 words
The voice recognition system as claimed in claim 1 wherein all the comparison vectors (y') are adapted to the present second spectral vector (n).
Claim 8
Dependent Claim
74 words
The voice recognition system as claimed in claim 1 wherein one index (j) is assigned at least a reference vector (y) and one or various successive second spectral vectors (n(i), n(i-1), . . . ) to form a respective number of comparison vectors (y') for this index (j); and for a comparison with the first spectral vectors (x), the individual comparison vectors (y') are weighted by adding weights (g) to the respective vector distance.
Claim 9
Dependent Claim
44 words
The voice recognition system as claimed in claim 1, wherein said voice recognition system recognizes speech signals of a near-end user transmitted along an acoustic path to a telecommunication device and controls sub-functions of the telecommunication device in accordance with the recognized speech signals.
Claim 10
Dependent Claim
33 words
The voice recognition system as claimed in claim 1, wherein said voice recognition system recognizes speech signals transmitted by a telephone network to a telecommunication device and controls sub-functions of the telecommunication device.
Claim 11
Dependent Claim
19 words
The voice recognition system as claimed in claim 8 wherein a mobile telephone is used as the telecommunication device.
Claim 12
Dependent Claim
20 words
The voice recognition system as claimed in claim 8 wherein a telephone answering machine is used as the telecommunication device.
Claim 13
Dependent Claim
24 words
The voice recognition system as claimed in claim 1, wherein said voice recognition system controls sub-functions of a radio receiver of a motor vehicle.
Claim 14
Dependent Claim
20 words
The voice recognition system as claimed in claim 1, wherein said voice recognition system controls sub-functions of a motor vehicle.
Claim 15
Dependent Claim
36 words
An electric device comprising a voice recognition system as claimed in claims 1, wherein sub-functions of this device or of another electric device are controlled in dependence on recognized speech signals of the voice recognition system.
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) | 2002-01-25 | 2002-07-27 | 2003-01-26 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2006-01-25 | 2006-07-27 | 2007-01-26 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2010-01-25 | 2010-07-27 | 2011-01-26 | $7700.00 | Window Closed / Lapsed |