Speech transcription using multiple data sources
This disclosure describes transcribing speech using audio, image, and other data. A system is described that includes an audio capture system configured to capture audio data associated with a plurality of speakers, an image capture system configured to capture images of one or more of the plurality of speakers, and a speech processing engine. The speech processing engine may be configured to recognize a plurality of speech segments in the audio data, identify, for each speech segment of the plurality of speech segments and based on the images, a speaker associated with the speech segment, transcribe each of the plurality of speech segments to produce a transcription of the plurality of speech segments including, for each speech segment in the plurality of speech segments, an indication of the speaker associated with the speech segment, and analyze the transcription to produce additional data derived from the transcription.
1. A system comprising:
an audio capture system configured to capture audio data associated with a plurality of speakers;
an image capture system configured to capture images of one or more of the plurality of speakers; and
a speech processing engine configured to:
recognize a plurality of speech segments in the audio data,
identify, for each speech segment of the plurality of speech segments and based on the images, a speaker associated with the speech segment,
transcribe each of the plurality of speech segments to produce a transcription of the plurality of speech segments including, for each speech segment in the plurality of speech segments, an indication of which speaker is associated with the speech segment, and
analyze the transcription to identify an instance in which one of the speakers asks another one of the speakers to perform a task, and based on the identified instance, generate a list of tasks.
2. The system of claim 1 , wherein to identify the speaker associated with the speech segment, the speech processing engine is further configured to:
identify the speaker based on external data about the speaker.
3. The system of claim 2 , wherein the external data about the speaker includes:
information about a calendar maintained by the speaker.
4. The system of claim 2 , wherein the external data about the speaker includes:
information about a location of the speaker.
5. The system of claim 1 , wherein the speech processing engine is further configured to generate additional data that includes:
an audio stream including a modified version of the speech segments associated with at least one of the plurality of speakers.
6. The system of claim 1 , wherein the speech processing engine is further configured to generate additional data that includes at least one of:
statistics about the transcription including number of words spoken by the speaker, tone of the speaker, information about filler words used by the speaker, percent of time the speaker spoke, information about profanity used, information about the length of words used, a summary of the transcription, or sentiment of the speaker.
7. The system of claim 1 , wherein to recognize the plurality of speech segments, the speech processing engine is further configured to:
recognize, based on the images, the plurality of speech segments.
8. The system of claim 1 , wherein to identify the speaker, the speech processing engine is further configured to:
detect one or more faces in the images.
9. The system of claim 8 , wherein to detect the one or more faces, the speech processing engine is further configured to:
detect one or more faces in the images with moving lips.
10. A method comprising:
capturing audio data associated with a plurality of speakers;
capturing images of one or more of the plurality of speakers;
recognizing a plurality of speech segments in the audio data;
identifying, for each speech segment of the plurality of speech segments and based on the images, a speaker associated with the speech segment;
transcribing each of the plurality of speech segments to produce a transcription of the plurality of speech segments including, for each speech segment in the plurality of speech segments, an indication of the speaker associated with the speech segment; and
analyzing the transcription to identify an instance in which one of the speakers asks another one of the speakers to meet later, and based on the identified instance, generating a calendar invitation.
11. The method of claim 10 , wherein identifying the speaker associated with the speech segment includes:
identifying the speaker based on external data about the speaker.
12. The method of claim 11 , wherein the external data about the speaker includes:
information about a calendar maintained by the speaker.
13. The method of claim 11 , wherein the external data about the speaker includes:
information about a location of the speaker.
14. The method of claim 10 , further comprising:
generating additional data that includes: an audio stream including a modified version of the speech segments associated with at least one of the plurality of speakers.
15. The method of claim 10 , further comprising:
generating additional data, wherein the additional data includes at least one of: statistics about the transcription including number of words spoken by the speaker, tone of the speaker, information about filler words used by the speaker, percent of time the speaker spoke, information about profanity used, information about the length of words used, a summary of the transcription, or sentiment of the speaker.
16. The method of claim 10 , wherein recognizing the plurality of speech segments includes:
recognizing, based on the images, the plurality of speech segments.
17. The method of claim 10 , wherein identifying the speaker includes:
detecting one or more faces in the images.
18. The method of claim 17 , wherein detecting the one or more faces includes:
detecting one or more faces in the images with moving lips.
19. A non-transitory computer-readable storage medium comprising instructions that, when executed, configure processing circuitry of a computing system to:
capture audio data associated with a plurality of speakers;
capture images of one or more of the plurality of speakers;
recognize a plurality of speech segments in the audio data,
identify, for each speech segment of the plurality of speech segments and based on the images, a speaker associated with the speech segment,
transcribe each of the plurality of speech segments to produce a transcription of the plurality of speech segments including, for each speech segment in the plurality of speech segments, an indication of which speaker is associated with the speech segment, and
analyze the transcription to identify an instance in which one of the speakers asks another one of the speakers to perform a task, and based on the identified instance, generate a list of tasks.
20. The non-transitory computer-readable storage medium of claim 19 , wherein the instructions that configure the processing circuitry to identify the speaker further include instructions that configure the processing circuitry to:
identify the speaker based on external data about the speaker.