Researchers at India’s International Institute of Information Technology (IIIT) developed software that automatically translates a video of a person speaking in one language into another language, while matching their lip movements to the new words.
The LipGAN algorithm is actually several algorithms functioning in concert, with one recognizing the words being spoken, and another translating the words into a target language. A text-to-speech algorithm produces the new sounds, while yet another algorithm animates synchronizes lip movements and facial movements with words spoken in the new language.
IIIT's Prajwal Renukanand suggested the technology could help translate TV shows or movies for international audiences.
From New Scientist
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA