A system for voicing sign language was developed in Russia

No time to read?
Get a summary

Novosibirsk State Technical University (NSTU NETI) has developed a system for recognizing and voicing Russian sign language. In this respect reports The Interfax agency cites Mikhail Grif, professor of the NSTU NETI Department of Automatic Control Systems.

The system recognizes movements in a continuous stream using a camera such as a smartphone. The service, which is a mobile application, can reversely translate human speech into sign language. The emergence of such a two-way translator is expected in late 2024 or early 2025.

Grif explained that in order for the system to work properly, the software must recognize approximately 5 thousand movements with 90 percent accuracy. Currently, the development of NSTU NETI has trained only 400 movements.

“Currently we can safely recognize 400 movements, this is of course not enough. We counted 8.5 thousand movements in various dictionaries, but there are probably even more. We estimate that the maximum number of gestures will be around 10 thousand,” Grief said.

He also added that one of the biggest problems when translating backwards (from voice to gestures) is the ambiguity of some words and synonyms.

Previously in Russia developed A new protection system for drones.

No time to read?
Get a summary
Previous Article

Prime Minister Morawiecki: Long live Poland. Free, democratic and independent. Simply – ours, the only one, the most beautiful and most common

Next Article

‘MasterChef Celebrity’ postpones its semi-final due to this spectacular event on TVE