Systems and methods for speech signal processing to transcribe speech

Patent metadata

Field	Value
Publication number	US-11107482-B2
Application number	US-201916704561-A
Country	US
Kind code	B2
Filing date	Dec 5, 2019
Priority date	Feb 28, 2018
Publication date	Aug 31, 2021
Grant date	Aug 31, 2021

Abstract

The present disclosure relates to systems and methods for speech signal processing on a signal to transcribe speech. In one implementation, the system may include a memory storing instructions and a processor configured to execute the instructions. The instructions may include instructions to receive the signal, determine if at least a portion of data in the signal is missing, and when at least a portion of data is missing: process the signal using a hidden Markov model to generate an output; using the output, calculate a set of possible contents to fill a gap due to the missing data portion, with each possible content having an associated probability; based on the associated probabilities, select one of the set of possible contents; and using the selected possible content, update the signal.

First claim

What is claimed is:

1. A system for speech signal processing on a signal to transcribe speech, the system comprising:
a memory storing instructions; and
a processor configured to execute the instructions to:
determine that at least a portion of data in the signal is missing,
process the signal using one or more neural networks to generate a plurality of transition probabilities,
process the signal using a hidden Markov model applying the plurality of transition probabilities to generate an output,
using the output, calculate a set of possible contents to fill a gap due to the missing data portion, and
update the signal using at least one of the set of possible contents.

2. The system of claim 1 , wherein calculating the set of possible contents comprises using a database indexing waveforms to at least a portion of the output.

3. The system of claim 2 , wherein the database is constructed from one or more training sets.

4. The system of claim 3 , wherein the processor is further configured to update the database based on the updated signal.

5. The system of claim 4 , wherein the update to the database is further based on feedback received from one or more users.

6. The system of claim 4 , wherein updating the database comprises reducing a loss function associated with the database.

7. The system of claim 1 , wherein the signal is further processed using one or more neural networks.

8. The system of claim 7 , wherein calculating the set of possible contents comprises using the output of the hidden Markov model as the set of possible contents and obtaining associated probabilities for the set of possible contents using the one or more neural networks.

9. The system of claim 1 , wherein the processor is further configured to send the updated signal to one or more participants in a communications session.

10. The system of claim 9 , wherein the communications session comprises at least one of a video conference and an audio conference.

11. A computer-implemented method for speech signal processing on a signal to transcribe speech, the method comprising:
determining that at least a portion of data in the signal is missing;
processing the signal using one or more neural networks to generate a plurality of transition probabilities;
processing the signal using a hidden Markov model applying the plurality of transition probabilities to generate an output;
using the output, calculating a set of possible contents to fill a gap due to the missing data portion; and
updating the signal using at least one of the set of possible contents.

12. The method of claim 11 , wherein calculating the set of possible contents comprises using a database of indexing waveforms to at least a portion of the output.

13. The method of claim 12 , further comprising updating the database based on the updated signal.

14. The method of claim 13 , wherein updating the database comprises reducing a loss function associated with the database.

15. The method of claim 11 , wherein the signal is further processed using one or more neural networks.

16. The method of claim 15 , wherein calculating the set of possible contents comprises using probabilities output by the one or more neural networks to calculate the set of possible contents using the hidden Markov model.

17. The method of claim 11 , further comprising sending the updated signal to one or more participants in at least one of a video conference or an audio conference.

18. A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause the at least one processor to:
determine that at least a portion of data in the signal is missing;
process the signal using one or more neural networks to generate a plurality of transition probabilities;
process the signal using a hidden Markov model applying the plurality of transition probabilities to generate an output;
using the output, calculate a set of possible contents to fill a gap due to the missing data portion; and
update the signal using at least one of the set of possible contents.

19. The non-transitory medium of claim 18 , wherein calculating the set of possible contents comprises using a database of indexing waveforms to at least a portion of the output.

20. The non-transitory medium of claim 18 , wherein calculating the set of possible contents comprises using the output of the hidden Markov model as the set of possible contents and obtaining associated probabilities for the set of possible contents using the one or more neural networks.

Assignees

• Ringcentral Inc

Inventors

• Li Xiaoming
• Khan Ehtesham
• Panattu Sethumadhavan Santosh

Classifications

• G06N3/047

  Probabilistic or stochastic networks · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/09

  Supervised learning · CPC title

• G10L15/26 Primary

  Speech to text systems (G10L15/08 takes precedence) · CPC title