 |
| |
The company Intel has released new software that allows engineers to create computer software capable of reading the user's lips, which can improve speech recognition applications, according to English news agency EFE.
The largest chipmaker in the world reported the release, which seeks to address weaknesses of voice recognition software, usually affected by background noise or poor tuning of the microphone.
Software Audio Visual Speech Recognition (AVSR) allows, according to Intel, the computer detected a speaker's face and mouth movements.
Thus, synchronization of data obtained through the identification of the video with voice recognition allows much more accurate.
Most AVSR software team is in the Intel Research Center in Beijing, China, employing more than forty research scientists and engineers in computer science.
0 comments:
Post a Comment