Vera Voice

2019: Announcement

At the end of October, 2019 it was announced start of Vera Voice service which allows to synthesize voices of celebrities in Russian. It is a joint product of the company Screenlife Technologies of Timur Bekmambetov and developer HR-Robot Vera service Stafory.

Vera Voice is the cornerstone the neuronet capable to learn to copy the speech of people from original records. Development will be used in interactive mobile applications based on movies and series in which users will be able "to talk" to characters.

It was announced start of Vera Voice service which allows to synthesize voices of celebrities in Russian

Also service can be used for a postscoring of audiobooks, advertizing, computer games and other thematic products.

According to the cofounder of Stafory Vladimir Sveshnikov, for synthesis of a voice there are enough several hours of the audiodata read by the owner of a voice, and the digitized text which he reads. The audio stream and the text are unloaded in several neural networks at the same time, one generates the spectrogram of a sound (its visual image), and the second — directly a sound, he told.

Screenlife Technologies and Stafory are going to sign agreements with owners of a voice — Vera Voice will become the agent on synthesis of his speech. The first pilot should start in November, 2019.

The agreement on legal use of a voice of the co-founder of Cube Cubed studio Ruslan Gabidullin for record of audiobooks, interactive advertizing, games and voice cards was already signed.

Also Vera Voice conducts negotiations on use of its technologies with the agency of audio-communications Unisound, the first projects can appear at the beginning of 2020.

Authors of the project intend to protect the interests of owners of voices and to fight against fakes. According to Timur Bekmambetov, "catch counterfeits and protect the famous people from unauthorized copying of their images — one of the main tasks".^[1]

You look also (voice assistants)

Notes