Distributed user input to text generated by a speech to text transcription service
AssigneeMicrosoft Corporation
InventorsJeremy Edward Cath, Timothy Edwin Harris, Marc Mercuri, James Oliver Tisdale, III
Life of this patent
Estimating the term…
Abstract
A particular method includes receiving, at a representational state transfer endpoint device, a first user input related to a first speech to text conversion performed by a speech to text transcription service. The method also includes receiving, at the representational state transfer endpoint device, a second user input related to a second speech to text conversion performed by the speech to text transcription service. The method includes processing of the first user input and the second user input at the representational state transfer endpoint device to generate speech to text adjustment information.
Claims20
A method comprising:
- receiving, at a computer device configured to process user input related to conversion of speech data to text, a first user input from a first client device related to a first speech to text conversion performed by a speech to text transcription service, wherein the first user input includes a first user rating and a phoneme alternative, wherein the speech to text transcription service is configured to receive first speech data from the first client device and to perform speech to text transcription on the first speech data to generate first text and to generate a first payload that includes the first text and first attributes, and wherein the first attributes include a user feedback scale range and an identifier of the computer device;
- receiving, at the computer device, a second user input from a second client device related to a second speech to text conversion performed by the speech to text transcription service;
- and
- processing the first user input and the second user input at the computer device to generate speech to text adjustment information.
A computer readable storage device including instructions executable by a computer to:
- receive captured speech data from a plurality of client devices;
- convert the captured speech data to text for each of the plurality of client devices;
- and
- send payload data that includes the text to each of the plurality of client devices, wherein the payload data includes at least one attribute that indicates a location of a remote computer device configured to process user input related to accuracy of the conversion of the captured speech data to the text, and wherein the payload data is configured to enable each client device to collect user input and send the collected user input to the remote computer device for processing, the collected user input including a vote, source language information, and user location information, wherein the payload data is also configured to enable the collected user input to be used to perform a rating analysis to generate speech to text adjustment information.
A computer implemented method comprising:
- sending captured speech data to a speech to text transcription service;
- receiving a payload including text from the speech to text transcription service;
- displaying the text at a display of an electronic device;
- displaying a user interface at the display, wherein the user interface enables receipt of user input regarding accuracy of the text;
- receiving the user input and at least one phoneme alternative to the text;
- and
- communicating the user input and the at least one phoneme alternative to a remote device, wherein the user input communicated to the remote device is configured to be useable with user input communicated to the remote device by a plurality of client devices, wherein each specific user input received from each specific client device is generated by capturing speech data at the specific client device, sending the speech data to the speech to text transcription service, and receiving a specific payload at the specific client device, wherein the specific payload includes attributes including a user feedback scale range, at least one phoneme alternative to transcribed text associated with the speech data, and an identifier of the remote device.
Dependencies are inferred from the claim text.