US 12,437,766 B2
Autocorrection of pronunciations of keywords in audio/videoconferences
Daina Emmanuel, Bangalore (IN); and Padmassri Chandrashekar, Karnataka (IN)
Assigned to Adeia Guides Inc., San Jose, CA (US)
Filed by Adeia Guides Inc., San Jose, CA (US)
Filed on Jun. 28, 2023, as Appl. No. 18/215,580.
Application 18/215,580 is a continuation of application No. 17/363,824, filed on Jun. 30, 2021, granted, now 11,727,940.
Prior Publication US 2023/0343343 A1, Oct. 26, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 17/00 (2013.01); G10L 17/14 (2013.01); G10L 17/22 (2013.01); G10L 19/16 (2013.01); G10L 25/57 (2013.01); H04L 65/403 (2022.01)
CPC G10L 17/14 (2013.01) [G10L 17/22 (2013.01); G10L 19/167 (2013.01); G10L 25/57 (2013.01); H04L 65/403 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
receiving audio data from a conference;
performing speech analysis and speech recognition on the audio data;
based on the speech analysis and the speech recognition, determining an indication of a mispronounced keyword that forms a part of the audio data;
replacing the portion of the audio data having the indication of the mispronounced keyword, in substantially real time, with a corrected portion of the audio data by referencing a database of keywords; and
transmitting an audio playlist comprising one or more bitrate variants as a corrected or re-encoded audio playlist.
 
10. A system comprising:
circuitry configured to:
receive audio data from a conference;
perform speech analysis and speech recognition on the audio data;
based on the speech analysis and the speech recognition, determine an indication of a mispronounced keyword that forms a part of the audio data;
replace the portion of the audio data having the indication of the mispronounced keyword, in substantially real time, with a corrected portion of the audio data by referencing a database of keywords; and
transmit an audio playlist comprising one or more bitrate variants as a corrected or re-encoded audio playlist.