Switching between speech recognition systems
A method may include obtaining first audio data originating at a first device during a communication session between the first device and a second device. The method may also include obtaining an availability of revoiced transcription units in a transcription system and in response to establishment of the communication session, selecting, based on the availability of revoiced transcription units, a revoiced transcription unit instead of a non-revoiced transcription unit to generate a transcript of the first audio data. The method may also include obtaining revoiced audio generated by a revoicing of the first audio data by a captioning assistant and generating a transcription of the revoiced audio using an automatic speech recognition system. The method may further include in response to selecting the revoiced transcription unit, directing the transcription of the revoiced audio to the second device as the transcript of the first audio data.
1. A method comprising:
obtaining first audio data originating at a first device during a communication session between the first device and a second device, the communication session configured for verbal communication;
obtaining an availability of revoiced transcription units in a transcription system;
in response to establishment of the communication session, selecting, based on the availability of revoiced transcription units, a revoiced transcription unit instead of a non-revoiced transcription unit to generate a transcript of the first audio data to direct to the second device such that only the transcript of the first audio data generated by the selected revoiced transcription unit is directed to the second device during the communication session and the non-selected non-revoiced transcription unit does not generate a transcript of the first audio data of the communication session;
obtaining, by the revoiced transcription unit, revoiced audio generated by a revoicing of the first audio data by a captioning assistant;
generating, by the revoiced transcription unit, a transcription of the revoiced audio using an automatic speech recognition system;
in response to selecting the revoiced transcription unit, directing the transcription of the revoiced audio to the second device as the transcript of the first audio data;
obtaining second audio data originating at a third device during a second communication session between the third device and the second device;
obtaining a second availability of the revoiced transcription units in the transcription system;
in response to establishment of the second communication session, selecting, based on the second availability of the revoiced transcription units, the non-revoiced transcription unit instead of one of the revoiced transcription units to generate a transcript of the second audio data to direct to the second device such that none of the revoiced transcription units generate a transcript of the second audio data; and
generating, by the selected non-revoiced transcription unit, a transcription of the second audio data.
2. The method of claim 1 , wherein the availability of revoiced transcription units is based on one or more of: a current peak number of transcriptions being generated, a current average number of transcriptions being generated, a projected peak number of transcriptions to be generated, a projected average number of transcriptions to be generated, a projected number of revoiced transcription units, and a number of available revoiced transcription units.
3. The method of claim 1 , wherein the availability of revoiced transcription units is based on four or more of: a current peak number of transcriptions being generated, a current average number of transcriptions being generated, a projected peak number of transcriptions to be generated, a projected average number of transcriptions to be generated, a projected number of revoiced transcription units, and a number of available revoiced transcription units.
4. The method of claim 1 , wherein the automatic speech recognition system is trained specifically for speech of the captioning assistant.
5. The method of claim 1 , wherein the automatic speech recognition system is adapted for speech of the captioning assistant and a second automatic speech recognition system used by the non-revoiced transcription unit is trained for a plurality of speakers.
6. The method of claim 1 , wherein the availability of revoiced transcription units is based on a current peak number of transcriptions being generated, a current average number of transcriptions being generated, a projected peak number of transcriptions to be generated, a projected average number of transcriptions to be generated, a projected number of revoiced transcription units, and a number of available revoiced transcription units.
7. The method of claim 1 , wherein the transcription system includes the non-revoiced transcription unit and one or more additional non-revoiced transcription units.
8. At least one non-transitory computer-readable media configured to store one or more instructions that in response to being executed by at least one computing system cause performance of the method of claim 1 .
9. A system comprising:
at least one computing system;
at least one computer-readable media coupled to the at least one computing system, the at least one computer-readable media configured to store one or more instructions that in response to being executed by the at least one computing system cause performance of operations, the operations comprising:
obtain first audio data originating at a first device during a communication session between the first device and a second device, the communication session configured for verbal communication;
obtain an availability of revoiced transcription units in a transcription system;
in response to establishment of the communication session, select, based on the availability of revoiced transcription units in the transcription system, a revoiced transcription unit instead of a non-revoiced transcription unit to generate a transcript of the first audio data that is directed to the second device and no non-selected non-revoiced transcription unit generates a transcript of the first audio data of the communication session;
in response to selecting the revoiced transcription unit, obtain a transcription of revoiced audio generated using an automatic speech recognition system of the revoiced transcription unit, the revoiced audio generated by a revoicing of the first audio data by a captioning assistant associated with the revoiced transcription unit;
direct the transcription of the revoiced audio to the second device as the transcript of the first audio data;
obtain second audio data originating at a third device during a second communication session between the third device and the second device;
obtain a second availability of the revoiced transcription units in the transcription system;
in response to establishment of the second communication session, select, based on the second availability of the revoiced transcription units, the non-revoiced transcription unit instead of one of the revoiced transcription units to generate a transcript of the second audio data to direct to the second device such that none of the revoiced transcription units generate a transcript of the second audio data; and
obtain a transcription of the second audio data generated by the selected non- revoiced transcription unit.
10. The system of claim 9 , wherein the availability of revoiced transcription units is based on one or more of: a current peak number of transcriptions being generated, a current average number of transcriptions being generated, a projected peak number of transcriptions to be generated, a projected average number of transcriptions to be generated, a projected number of revoiced transcription units, and a number of available revoiced transcription units.
11. The system of claim 9 , wherein the availability of revoiced transcription units is based on four or more of: a current peak number of transcriptions being generated, a current average number of transcriptions being generated, a projected peak number of transcriptions to be generated, a projected average number of transcriptions to be generated, a projected number of revoiced transcription units, and a number of available revoiced transcription units.
12. The system of claim 9 , wherein the automatic speech recognition system is adapted for speech of the captioning assistant.
13. The system of claim 9 , wherein the automatic speech recognition system is trained specifically for speech of the captioning assistant and a second automatic speech recognition system used by the non-revoiced transcription unit is trained for a plurality of speakers.
14. The system of claim 9 , wherein the availability of revoiced transcription units is based on a current peak number of transcriptions being generated, a current average number of transcriptions being generated, a projected peak number of transcriptions to be generated, a projected average number of transcriptions to be generated, a projected number of revoiced transcription units, and a number of available revoiced transcription units.
15. The system of claim 9 , wherein the system is included in the transcription system.
16. The system of claim 9 , wherein the transcription system includes the non-revoiced transcription unit and one or more additional non-revoiced transcription units.
17. A method comprising:
obtaining first audio data originating at a first device during a communication session between the first device and a second device;
obtaining an availability of revoiced transcription units in a transcription system;
in response to establishment of the communication session, selecting, based on the availability of revoiced transcription units in the transcription system, a revoiced transcription unit instead of a non-revoiced transcription unit to generate a transcript of the first audio data that is directed to the second device and the non-selected non-revoiced transcription unit does not generate a transcript of the first audio data of the communication session;
in response to selecting the revoiced transcription unit, obtaining a transcription of revoiced audio generated using an automatic speech recognition system of the revoiced transcription unit, the revoiced audio generated by a revoicing of the first audio data;
directing the transcription of the revoiced audio to the second device as the transcript of the first audio data;
obtaining second audio data originating at a third device during a second communication session between the third device and the second device;
obtaining a second availability of the revoiced transcription units in the transcription system;
in response to establishment of the second communication session, selecting, based on the second availability of the revoiced transcription units, the non-revoiced transcription unit instead of one of the revoiced transcription units to generate a transcript of the second audio data to direct to the second device such that none of the revoiced transcription units generate a transcript of the second audio data; and
generating, by the selected non-revoiced transcription unit, a transcription of the second audio data.
18. At least one non-transitory computer-readable media configured to store one or more instructions that in response to being executed by at least one computing system cause performance of the method of claim 17 .
19. The method of claim 17 , wherein the transcription system includes the non-revoiced transcription unit and one or more additional non-revoiced transcription units.