Voice recognition with parallel gender and age normalization
First Claim
1. A method for voice recognition, the method comprising:
- obtaining a voice signal;
performing two or more voice recognition analyses in parallel on the voice signal, wherein each voice recognition analysis uses a filter bank defined by a different maximum frequency and a different minimum frequency and wherein each voice recognition analysis produces a recognition probability ri of recognition of one or more speech units, whereby there are two or more recognition probabilities ri, wherein each recognition probability ri represents a probability that the voice signal matches a recognized unit of speech, wherein performing the two or more voice recognition analyses in parallel includes performing a first voice recognition analysis on the voice signal based on a first maximum frequency fmax and a first minimum frequency fmin;
wherein fmin and fmax are adjusted dynamically during the first voice recognition; and
performing one or more additional voice recognition analyses on the voice signal based on a different maximum frequency given by fmax±
Δ
fmax and a different minimum frequency given by fmin±
Δ
fmin,where Δ
fmax<
fmaxand Δ
fmin<
fmin; and
determining a final recognition probability Pf based on the two or more recognition probabilities ri;
determining a confidence measure based on the final recognition probability and the two or more recognition probabilities ri; and
using the confidence measure as an acceptance criterion for the final recognition probability Pf after determining the final recognition probability Pf.
4 Assignments
0 Petitions
Accused Products
Abstract
Methods and apparatus for voice recognition are disclosed. A voice signal is obtained and two or more voice recognition analyses are performed on the voice signal. Each voice recognition analysis uses a filter bank defined by a different maximum frequency and a different minimum frequency and wherein each voice recognition analysis produces a recognition probability ri of recognition of one or more speech units, whereby there are two or more recognition probabilities ri. The maximum frequency and the minimum frequency may be adjusted every time speech is windowed and analyzed. A final recognition probability Pf is determined based on the two or more recognition probabilities ri.
-
Citations
23 Claims
-
1. A method for voice recognition, the method comprising:
-
obtaining a voice signal; performing two or more voice recognition analyses in parallel on the voice signal, wherein each voice recognition analysis uses a filter bank defined by a different maximum frequency and a different minimum frequency and wherein each voice recognition analysis produces a recognition probability ri of recognition of one or more speech units, whereby there are two or more recognition probabilities ri, wherein each recognition probability ri represents a probability that the voice signal matches a recognized unit of speech, wherein performing the two or more voice recognition analyses in parallel includes performing a first voice recognition analysis on the voice signal based on a first maximum frequency fmax and a first minimum frequency fmin;
wherein fmin and fmax are adjusted dynamically during the first voice recognition; andperforming one or more additional voice recognition analyses on the voice signal based on a different maximum frequency given by fmax±
Δ
fmax and a different minimum frequency given by fmin±
Δ
fmin,where Δ
fmax<
fmaxand Δ
fmin<
fmin; anddetermining a final recognition probability Pf based on the two or more recognition probabilities ri; determining a confidence measure based on the final recognition probability and the two or more recognition probabilities ri; and using the confidence measure as an acceptance criterion for the final recognition probability Pf after determining the final recognition probability Pf. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 23)
-
-
18. A voice recognition system, comprising:
-
an interface adapted to obtain a voice signal; one or more processors coupled to the interface; and a memory coupled to the interface and the processor, the memory having embodied therein a set of processor readable instructions for configured to implement a method for voice recognition, the processor readable instructions including; one or more instructions that direct the one or more processors to perform two or more voice recognition analyses in parallel on the voice signal, wherein each voice recognition analysis uses a filter bank defined by a different maximum frequency and a different minimum frequency and wherein each voice recognition analysis produces a recognition probability ri , of recognition of one or more speech units, whereby there are two or more recognition probabilities ri, wherein each recognition probability ri represents a probability that the voice signal matches a recognized unit of speech, wherein each recognition probability ri represents a probability that the voice signal matches a recognized unit of speech, wherein performing the two or more voice recognition analyses in parallel includes performing a first voice recognition analysis on the voice signal based on a first maximum frequency f max and a first minimum frequency fmin;
wherein fmin and fmax are adjusted dynamically during the first voice recognition; andperforming one or more additional voice recognition analyses on the voice signal based on a different maximum frequency given by fmax±
Δ
fmax and a different minimum frequency given by fmin±
Δ
fmin,where Δ
fmax <
fmax and Δ
fmin<
fmin; andan instruction to determine a final recognition probability Pf based on the two or more recognition probabilities; and an instruction to determine a confidence measure based on the final recognition probability and the two or more recognition probabilities ri; and an instruction to use the confidence measure as an acceptance criterion for the final recognition probability Pf after determining the final recognition probability Pf. - View Dependent Claims (19, 20, 21, 22)
-
Specification