Patent · US Active

End-to-end neural networks for speech recognition and classification

US10720151B2 · kind B2 · utility

5Cited by

34References

20Claims

0Family size

Assignee

Deepgram, Inc. · US

Inventors

Adam Sypniewski · Ypsilanti, US
Jeff Ward · San Francisco, US
Scott Stephenson · Burlingame, US

Key dates

Filing date	Aug 22, 2018
Grant date	Jul 21, 2020
Priority date	—
Expiry date	Aug 22, 2038

Classification

Technology area (CPC G)Physics
CPC primaryG10L2015/081
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.