The technology segmentation of audio and speech recognition, developed by a group of companies of the MDGs, is recognized as the best at the international competition CHiME Speech Separation and Recognition Challenge (CHiME-6). On Thursday, May 7, reported the press service of the developer.”To CHiME-5 contestants were decided by the so-called cocktail party problem — recognition of spontaneous speech of several speakers in the conditions of partial overlap of speech and noise, that is, in a typical situation of communication at the party. This unit is required to operate the segmented (already allocated) speech. The novelty and peculiarity of the CHiME-6 was the fact that for the first time in history, the contestants were asked to solve the same problem, but working with non-segmented speech, at the same time — overlapping speech to 20%,” — said in a press release.
Entries for the competition were made on 20 nights in these houses, where people freely talked, joked, laughed, cooked, ate, and washed dishes.
The organizers put the participant’s goal is to create a recognition system that “listen” to the recording, and gives you a full transcript with the least amount of errors. In the end, the winner was technology, developed by specialists of the MDGs.
“High-quality speech recognition of different speakers, with interrupting noise, allows you to display the services from the category of innovation in everyday use, improving the business and simplifying our lives,” — said General Director of a group of companies Dmitry Dyrmovsky the MDGs.
In early January the President of Russia Vladimir Putin in the annual message to the Federal Assembly said that the country is able to achieve a breakthrough in the development of artificial intelligence.
In November of last year, the Savings Bank has created the most powerful Russian supercomputer. This model helps to accelerate the development of services and processes based on artificial intelligence.