Method and apparatus for intelligent voice recognition
Methods and systems are described for recognizing, based on a voice input, a user and/or a voice command. An algorithm is described herein that processes data associated with a voice input. The data may indicate characteristics of the voice such as a gender, an age, or accent associated with the voice and other metadata. For example, the system may process the data and determine the gender of a voice. The determined characteristics may be used as an input into a voice recognition engine to improve the accuracy of identifying the user who spoke the voice input and identifying a voice command associated with the voice input. For example, the determined gender may be used as a parameter to improve the accuracy of an identified user (e.g., the speaker) or command. The algorithm may adjust, based on gender, parameters such as confidence thresholds used to match voices and voice commands.
1 . A device comprising:
one or more processors; and
memory storing instructions that, when executed by the one or more processors, cause the device to:
determine, based on data associated with a voice input, a gender associated with the voice input;
determine, based on the gender being associated with a female voice, a first confidence threshold;
determine, based on the voice input, a confidence value that the voice input was received from a user;
determine, based on the confidence value satisfying the first confidence threshold, to send an indication that the voice input was received from the user; and
send, based on determining to send the indication, the indication.
2 . The device of claim 1 , wherein the instructions, when executed by the one or more processors, further cause the device to:
determine, based on second data associated with a second voice input, a second gender associated with the second voice input;
determine, based on the second gender being associated with a male voice, a second confidence threshold, wherein the first confidence threshold is different than the second confidence threshold;
determine, based on the second voice input, a second confidence value that the second voice input was received from a second user;
determine, based on the second confidence value satisfying the second confidence threshold, to send a second indication that the second voice input was received from the second user; and
send, based on determining to send the second indication, the second indication.
3 . The device of claim 1 , wherein the first confidence threshold indicates a degree of accuracy needed to match the data with the user.
4 . The device of claim 1 , wherein the first confidence threshold minimizes at least one of: a false acceptance rate (FAR) associated with determining that the voice input was received from the user, or a false rejection rate (FRR) associated with determining that the voice input was received from the user.
5 . The device of claim 1 , wherein the first confidence threshold causes an equal error rate for a false acceptance rate (FAR) associated with determining that the voice input was received from the user and a false rejection rate (FRR) associated with determining that the voice input was received from the user.
6 . The device of claim 1 , wherein the determining the confidence value that the voice input was received from a user is further based on a comparison to stored information, wherein the stored information comprises an implicit user fingerprint comprising one or more features associated with the user.
7 . The device of claim 1 , wherein the instructions, when executed by the one or more processors, further cause the device to:
generate, based on a second voice input, an implicit user fingerprint comprising one or more features associated with the user, wherein the second voice input comprises at least one of: a previous voice input or an enrollment command comprising one or more words.
8 . The device of claim 1 , wherein the instructions, when executed by the one or more processors, further cause the device to:
determine, based at least in part on the first confidence threshold, a command.
9 . A non-transitory computer-readable medium storing instructions that, when executed, cause:
determining, based on data associated with a voice input, a gender associated with the voice input;
determining, based on the gender being associated with a female voice, a first confidence threshold;
determining, based on the voice input, a confidence value that the voice input was received from a user;
determining, based on the confidence value satisfying the first confidence threshold, to send an indication that the voice input was received from the user; and
sending, based on determining to send the indication, the indication.
10 . The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed, further cause:
determining, based on second data associated with a second voice input, a second gender associated with the second voice input;
determining, based on the second gender being associated with a male voice, a second confidence threshold, wherein the first confidence threshold is different than the second confidence threshold;
determining, based on the second voice input, a second confidence value that the second voice input was received from a second user;
determining, based on the second confidence value satisfying the second confidence threshold, to send a second indication that the second voice input was received from the second user; and
sending, based on determining to send the second indication, the second indication.
11 . The non-transitory computer-readable medium of claim 9 , wherein the first confidence threshold indicates a degree of accuracy needed to match the data with the user.
12 . The non-transitory computer-readable medium of claim 9 , wherein the first confidence threshold minimizes at least one of: a false acceptance rate (FAR) associated with determining that the voice input was received from the user, or a false rejection rate (FRR) associated with determining that the voice input was received from the user.
13 . The non-transitory computer-readable medium of claim 9 , wherein the first confidence threshold causes an equal error rate for a false acceptance rate (FAR) associated with determining that the voice input was received from the user and a false rejection rate (FRR) associated with determining that the voice input was received from the user.
14 . The non-transitory computer-readable medium of claim 9 , wherein the determining the confidence value that the voice input was received from a user is further based on a comparison to stored information, wherein the stored information comprises an implicit user fingerprint comprising one or more features associated with the user.
15 . The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed, further cause:
generating, based on a second voice input, an implicit user fingerprint comprising one or more features associated with the user, wherein the second voice input comprises at least one of: a previous voice input or an enrollment command comprising one or more words.
16 . The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed by the one or more processors, further cause the device to:
determining, based at least in part on the first confidence threshold, a command.
17 . A system comprising:
a first computing device configured to:
determine, based on data associated with a voice input, a gender associated with the voice input,
determine, based on the gender being associated with a female voice, a first confidence threshold,
determine, based on the voice input, a confidence value that the voice input was received from a user,
determine, based on the confidence value satisfying the first confidence threshold, to send an indication that the voice input was received from the user, and
send, based on determining to send the indication, the indication; and
a second computing device configured to:
receive the indication.
18 . The system of claim 17 , wherein the first computing device is further configured to:
determine, based on second data associated with a second voice input, a second gender associated with the second voice input;
determine, based on the second gender being associated with a male voice, a second confidence threshold, wherein the first confidence threshold is different than the second confidence threshold;
determine, based on the second voice input, a second confidence value that the second voice input was received from a second user;
determine, based on the second confidence value satisfying the second confidence threshold, to send a second indication that the second voice input was received from the second user; and
send, based on determining to send the second indication, the second indication.
19 . The system of claim 17 , wherein the first confidence threshold minimizes at least one of: a false acceptance rate (FAR) associated with determining that the voice input was received from the user, or a false rejection rate (FRR) associated with determining that the voice input was received from the user.
20 . The system of claim 17 , wherein the first confidence threshold causes an equal error rate for a false acceptance rate (FAR) associated with determining that the voice input was received from the user and a false rejection rate (FRR) associated with determining that the voice input was received from the user.