Identifying candidate passwords from captured audio

• US 8,898,064 B1
• Filed: 03/19/2012
• Issued: 11/25/2014
• Est. Priority Date: 03/19/2012
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Rawles, LLC (Amazon.com, Inc.)
• Inventors: Strom, Nikko, Thomas, Ryan P.
• Primary Examiner(s): PULLIAS, JESSE SCOTT
• Application Number:
  US13/424,104
• Time in Patent Office: 981 Days
• Field of Search: 704270-275
• US Class Current: 704/273
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06F 3/167 Audio in a user interface, e.g. using voice commands for navigating, audio feedback
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 2015/086 Recognition of spelled words

Abstract

A computing device configured to request a password from a user, capture audio after issuing the request, and determine a number of alternative candidate passwords most likely represented by the audio. After identifying the number of candidate passwords, the computing device may submit these candidate passwords, one at a time, to an entity until the entity grants the device access to an account associated with the user or until the device has submitted each candidate password. The account may comprise a network account (e.g., a wired or wireless network account), an online account (e.g., an email account, an account an online merchant, etc.), or the like.

Claims

1. An apparatus comprising;

a speaker;

a microphone;

a wireless network interface;

a processor; and

computer-readable media storing computer-executable instructions that, when executed on the processor, cause the processor to perform acts comprising;

causing the speaker to output a request for a password associated with a wireless network;

receiving audio captured by the microphone after the speaker outputs the request;

decoding the audio at least in part to identify alternative candidate passwords from the audio; and

submitting, one at a time, at least one of the alternative candidate passwords to a component of the wireless network until the wireless network interface successfully connects to the wireless network or until each of the alternative candidate passwords has been submitted;

wherein the decoding utilizes a language model, the acts further comprising;

identifying a candidate password of the alternative candidate passwords that, upon submittal to the component of the wireless network, results in the wireless network interface successfully connecting to the wireless network; and

sending the candidate password of the alternative candidate passwords over the wireless network and to an entity to allow the entity to train the language model or another language model using the candidate password of the alternative candidate passwords.

2. An apparatus as recited in claim 1, wherein the request instructs a user to spell the password associated with the wireless network.

3. An apparatus as recited in claim 1, wherein the submitting comprises submitting the alternative candidate passwords in accordance with a ranked list of the alternative candidate passwords.

4. An apparatus as recited in claim 1, wherein:

each of the alternative candidate passwords is associated with at least one of;

a probability or confidence level that the audio includes a sequence of sounds representing the respective candidate password, or a probability or confidence level that the audio includes a sequence of sounds representing a sequence of graphemes forming the respective candidate password; and

the submitting comprises submitting, one at a time, the alternative candidate passwords based at least in part on the respective probabilities or confidence levels of the alternative candidate passwords.

5. An apparatus as recited in claim 1, the acts further comprising causing the speaker to output an indication that a connection to the wireless network has been established.

6. An apparatus as recited in claim 1, the acts further comprising:

causing the speaker to output a request for an identifier of a wireless network;

receiving audio captured by the microphone after the speaker outputs the request for the identifier; and

identifying the wireless network at least partly based on the audio captured by the microphone after the speaker outputs the request for the identifier.

7. A method implemented at least in part by an electronic device coupled to a microphone, the method comprising:

determining, at least partly from audio captured by the microphone, a first candidate string of characters represented by an utterance from a user and a second candidate string of characters represented by the utterance from the user;

submitting the first candidate string of characters to a component of a network;

determining whether a connection has been established with the network after submitting the first candidate string of characters; and

submitting the second candidate string of characters to the component of the network at least partly in response to determining that the connection with the network has not been established after submitting the first candidate string of characters.

8. A method as recited in claim 7, wherein the first candidate string of characters is a most likely string of characters represented by the utterance of the user and the second candidate string of characters is a next most likely string of characters represented by the utterance of the user.

9. A method as recited in claim 7, wherein the submitting of the first candidate string of characters and the submitting of the second candidate string of characters comprises submitting each of the respective strings as a password for connecting with the network.

10. A method as recited in claim 7, further comprising instructing a user to spell a password for connecting to the network prior to determining the first candidate string of characters and the second candidate string of characters from the audio.

11. A method as recited in claim 7, wherein the determining further determines a predefined number of additional candidate strings of characters at least partly from the utterance from the user, and further comprising:

determining whether the connection has been established with the network after submitting the second candidate string of characters; and

submitting, one at a time, at least one of the additional candidate strings of characters to the component of the network.

12. Non-transitory computer-readable media storing computer-executable instructions that, when executed on a processor, cause the processor to perform acts comprising:

decoding audio at least in part to determine a number of alternative character strings from the audio; and

submitting, one at a time, the alternative character strings as a password until the earlier of authentication or until each of the number of alternative character strings has been submitted;

wherein the decoding utilizes a language model, the acts further comprising;

identifying a candidate character string of the alternative character strings that, upon submittal, results in authentication; and

sending the candidate character string of the alternative character strings over a wireless network and to an entity to allow the entity to train the language model or another language model using the candidate character string of the alternative character strings.

13. Non-transitory computer-readable media as recited in claim 12, wherein each of the alternative character strings is determined with reference to a probability or confidence level that the audio includes a series of sounds representing the respective character string.

14. Non-transitory computer-readable media as recited in claim 12, wherein each of the alternative character strings is determined with reference to a probability or confidence level that the audio includes a series of sounds representing a series of graphemes forming the respective character string.

15. Non-transitory computer-readable media as recited in claim 12, wherein the submitting comprises submitting the alternative character strings, one at a time, in an order based at least in part on respective probabilities or confidence levels that the audio includes respective series of sounds representing the respective character strings.

16. Non-transitory computer-readable media as recited in claim 12, wherein the submitting comprises submitting the alternative character strings, one at a time, in an order based at least in part on respective probabilities or confidence levels that the audio includes respective series of sounds representing respective series of graphemes forming the respective character strings.

17. Non-transitory computer-readable media as recited in claim 12, wherein the submitting comprises submitting the alternative character strings, one at a time, beginning with a character string corresponding to a series of sounds that is most likely to be included in the audio.

18. Non-transitory computer-readable media as recited in claim 12, wherein the submitting comprises submitting the alternative character strings, one at a time, beginning with a character string corresponding to a series of graphemes corresponding to a series of sounds that is most likely to be included in the audio.

19. Non-transitory computer-readable media as recited in claim 12, wherein the authentication is effective to grant access to an online account, a wireless network account, a wired network account, a computing device, an application, or a content item.

20. Non-transitory computer-readable media as recited in claim 12, the acts further comprising ranking the alternative character strings, and wherein the submitting of the alternative character strings comprises submitting the alternative character strings, one at a time, in accordance with the ranking.

21. Non-transitory computer-readable media as recited in claim 12, the acts further comprising outputting an indication of whether or not access to an account has been received.

22. Non-transitory computer-readable media as recited in claim 12, the acts further comprising outputting a request that a user utter the password for an account.

23. Non-transitory computer-readable media as recited in claim 12, the acts further comprising outputting a request that a user spell the password for an account.