Speech practice with media content synchronization
AssigneeInternational Business Machines Corporation
InventorsSudheesh S. Kairali, Sarbajit K. Rakshit
Life of this patent
Estimating the term…
Abstract
An embodiment includes detecting by a Speech Detection Component of a system a speech metric of a speaker in response to a reference speech. The embodiment includes responsive to the detected speech metric, computing by a Speech Analysis Component of the system a deviation metric between the speech metric and the reference speech. The embodiment includes training a machine learning model by a Speech Prediction Component of the system based on the deviation metric to generate a predicted speech pattern of the speaker. The embodiment also includes transforming by a Controller Component of the system the reference speech based on the predicted speech pattern.
Claims17
A computer-implemented method comprising:
- detecting by a system a speech metric of a speaker speaking a reference speech in response to a medium of the reference speech during an interaction between the speaker and the medium wherein the speech metric comprises tone, pronunciation, body movement and hand movement;
- responsive to detecting the speech metric, computing by the system a deviation metric between the speech metric and the reference speech;
- training a machine learning model by the system based on the deviation metric to output a trained machine learning model, the trained machine learning model generating a predicted speech pattern of the speaker;
- and
- transforming by the system the medium of the reference speech during the interaction between the speaker and the medium based on the predicted speech pattern wherein a processor, responsive to detecting a speech difficulty of the speaker during the interaction between the speaker and the medium based on the deviation metric between the speech metric and the reference speech, applies transformation to the medium of the reference speech during the interaction between the speaker and the medium to create a modified medium of the reference speech that responds to the speech difficulty of the speaker wherein the modified medium of the reference speech is further adjusted according to feedback of monitoring an output of the modified medium of the reference speech during the interaction between the speaker and the medium.
A computer program product comprising one or more computer readable storage media, and program instructions collectively stored on the one or more computer readable storage media, the program instructions executable by a processor to cause the processor to perform operations comprising:
- detecting by a system a speech metric of a speaker speaking a reference speech in response to a medium of the reference speech during an interaction between the speaker and the medium wherein the speech metric comprises tone, pronunciation, body movement and hand movement;
- responsive to detecting the speech metric, computing by the system a deviation metric between the speech metric and the reference speech;
- training a machine learning model by the system based on the deviation metric to output a trained machine learning model, the trained machine learning model generating a predicted speech pattern of the speaker;
- and
- transforming by the system the medium of the reference speech during the interaction between the speaker and the medium based on the predicted speech pattern wherein a processor, responsive to detecting a speech difficulty of the speaker during the interaction between the speaker and the medium based on the deviation metric between the speech metric and the reference speech, applies transformation to the medium of the reference speech during the interaction between the speaker and the medium to create a modified medium of the reference speech that responds to the speech difficulty of the speaker wherein the modified medium of the reference speech is further adjusted according to feedback of monitoring an output of the modified medium of the reference speech during the interaction between the speaker and the medium.
A computer system comprising a processor and one or more computer readable storage media, and program instructions collectively stored on the one or more computer readable storage media, the program instructions executable by the processor to cause the processor to perform operations comprising:
- detecting by a system a speech metric of a speaker speaking a reference speech in response to a medium of the reference speech during an interaction between the speaker and the medium wherein the speech metric comprises tone, pronunciation, body movement and hand movement;
- responsive to detecting the speech metric, computing by the system a deviation metric between the speech metric and the reference speech;
- training a machine learning model by the system based on the deviation metric to output a trained machine learning model, the trained machine learning model generating a predicted speech pattern of the speaker;
- and
- transforming by the system the medium of the reference speech during the interaction between the speaker and the medium based on the predicted speech pattern wherein a processor, responsive to detecting a speech difficulty of the speaker during the interaction between the speaker and the medium based on the deviation metric between the speech metric and the reference speech, applies transformation to the medium of the reference speech during the interaction between the speaker and the medium to create a modified medium of the reference speech that responds to the speech difficulty of the speaker wherein the modified medium of the reference speech is further adjusted according to feedback of monitoring an output of the modified medium of the reference speech during the interaction between the speaker and the medium.
Dependencies are inferred from the claim text.