US 12,307,214 B2
Hybrid language translation on mobile devices
Kee-Hyun Park, San Diego, CA (US); Sungrack Yun, Seongnam (KR); and Kyu Woong Hwang, Daejeon (KR)
Assigned to QUALCOMM Incorporated, San Diego, CA (US)
Filed by QUALCOMM Incorporated, San Diego, CA (US)
Filed on Sep. 23, 2022, as Appl. No. 17/952,025.
Prior Publication US 2024/0104311 A1, Mar. 28, 2024
Int. Cl. G06F 40/58 (2020.01); G06F 40/51 (2020.01); G10L 15/16 (2006.01)
CPC G06F 40/58 (2020.01) [G06F 40/51 (2020.01); G10L 15/16 (2013.01)] 24 Claims
OG exemplary drawing
 
1. A processor-implemented method for recognizing a natural language on a mobile device, comprising:
receiving, by a neural network on the mobile device, an audio input;
processing, by the neural network, the audio input to generate a translation of the audio input to local text representing the audio input;
generating a local confidence metric corresponding to a first accuracy of the local text representing the audio input;
transmitting, to a remote device, data corresponding to the audio input;
receiving, from the remote device, remote text representing the data and a remote confidence metric corresponding to a second accuracy of the remote text representing the data corresponding to the audio input; and
outputting, by the mobile device, one of the local text or the remote text based on a comparison of the local confidence metric and the remote confidence metric.