Patents by Inventor Keith P. Braho
Keith P. Braho has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20180310108Abstract: A system directs boom microphone placement. A microphone is configured to capture speech audio from a user and output corresponding electrical signals. A proximity sensor is situated adjacent the microphone and configured to produce output signals representative of a distance from the microphone to the user's face or mouth. A headset assembly includes a boom carrying the microphone and the proximity sensor, where the boom can be adjusted to a plurality of positions adjacent the user's face or mouth. Processing circuitry is coupled to receive the output signals from the proximity sensor and produce an output indicative that the microphone is outside a prescribed distance or range of distances from the user's face or mouth.Type: ApplicationFiled: April 21, 2017Publication date: October 25, 2018Inventors: Richard Sharbaugh, Ryan A. Zoschg, Keith P. Braho
-
Patent number: 9928829Abstract: Methods are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. Apparatus are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system.Type: GrantFiled: October 17, 2014Date of Patent: March 27, 2018Assignee: Vocollect, Inc.Inventors: Keith P. Braho, Jeffrey P. Pike, Lori A. Pike
-
Patent number: 9299344Abstract: Audio frames are classified as either speech, non-transient background noise, or transient noise events. Probabilities of speech or transient noise event, or other metrics may be calculated to indicate confidence in classification. Frames classified as speech or noise events are not used in updating models (e.g., spectral subtraction noise estimates, silence model, background energy estimates, signal-to-noise ratio) of non-transient background noise. Frame classification affects acceptance/rejection of recognition hypothesis. Classifications and other audio related information may be determined by circuitry in a headset, and sent (e.g., wirelessly) to a separate processor-based recognition device.Type: GrantFiled: July 1, 2015Date of Patent: March 29, 2016Assignee: Intermec IP Corp.Inventors: Keith P. Braho, David D. Hardek
-
Patent number: 9202458Abstract: A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. Apparatus are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The apparatus may further include a controller adapted to adjust an adaptation of the model for the word or various models for the various words, based on the error rate.Type: GrantFiled: October 11, 2010Date of Patent: December 1, 2015Assignee: Vocollect, Inc.Inventors: Keith P. Braho, Jeffrey P. Pike, Lori A. Pike
-
Publication number: 20150302853Abstract: Audio frames are classified as either speech, non-transient background noise, or transient noise events. Probabilities of speech or transient noise event, or other metrics may be calculated to indicate confidence in classification. Frames classified as speech or noise events are not used in updating models (e.g., spectral subtraction noise estimates, silence model, background energy estimates, signal-to-noise ratio) of non-transient background noise. Frame classification affects acceptance/rejection of recognition hypothesis. Classifications and other audio related information may be determined by circuitry in a headset, and sent (e.g., wirelessly) to a separate processor-based recognition device.Type: ApplicationFiled: July 1, 2015Publication date: October 22, 2015Inventors: Keith P. Braho, David D. Hardek
-
Patent number: 9076459Abstract: Audio frames are classified as either speech, non-transient background noise, or transient noise events. Probabilities of speech or transient noise event, or other metrics may be calculated to indicate confidence in classification. Frames classified as speech or noise events are not used in updating models (e.g., spectral subtraction noise estimates, silence model, background energy estimates, signal-to-noise ratio) of non-transient background noise. Frame classification affects acceptance/rejection of recognition hypothesis. Classifications and other audio related information may be determined by circuitry in a headset, and sent (e.g., wirelessly) to a separate processor-based recognition device.Type: GrantFiled: March 12, 2013Date of Patent: July 7, 2015Assignee: Intermec IP, Corp.Inventors: Keith P. Braho, David D. Hardek
-
Publication number: 20150039309Abstract: Methods are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. Apparatus are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system.Type: ApplicationFiled: October 17, 2014Publication date: February 5, 2015Inventors: Keith P. Braho, Jeffrey P. Pike, Lisa A. Pike
-
Patent number: 8868421Abstract: A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. Apparatus are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The apparatus may further include a controller adapted to adjust an adaptation of the model for the word or various models for the various words, based on the error rate.Type: GrantFiled: October 11, 2010Date of Patent: October 21, 2014Assignee: Vocollect, Inc.Inventors: Keith P. Braho, Jeffrey P. Pike, Lori A. Pike
-
Publication number: 20140278391Abstract: Audio frames are classified as either speech, non-transient background noise, or transient noise events. Probabilities of speech or transient noise event, or other metrics may be calculated to indicate confidence in classification. Frames classified as speech or noise events are not used in updating models (e.g., spectral subtraction noise estimates, silence model, background energy estimates, signal-to-noise ratio) of non-transient background noise. Frame classification affects acceptance/rejection of recognition hypothesis. Classifications and other audio related information may be determined by circuitry in a headset, and sent (e.g., wirelessly) to a separate processor-based recognition device.Type: ApplicationFiled: March 12, 2013Publication date: September 18, 2014Applicant: INTERMEC IP CORP.Inventors: Keith P. Braho, David D. Hardek
-
Publication number: 20110029312Abstract: Methods are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. Apparatus are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system.Type: ApplicationFiled: October 11, 2010Publication date: February 3, 2011Applicant: VOCOLLECT, INC.Inventors: Keith P. Braho, Jeffrey P. Pike, Lori A. Pike
-
Publication number: 20110029313Abstract: Methods are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. Apparatus are disclosed for identifying possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system.Type: ApplicationFiled: October 11, 2010Publication date: February 3, 2011Applicant: VOCOLLECT, INC.Inventors: Keith P. Braho, Jeffrey P. Pike, Lori A. Pike
-
Patent number: 7827032Abstract: Possible errors made by a speech recognition system may be identified without using a transcript of words input to the system. A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. An apparatus identifies possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system.Type: GrantFiled: October 6, 2006Date of Patent: November 2, 2010Assignee: Vocollect, Inc.Inventors: Keith P. Braho, Jeffrey P. Pike, Lori A. Pike