Specialists of the Chinese laboratory Baidu Research, owned by the largest search giant Celestial, created the algorithm of AI Deep Voice, capable of converting text to speech. Similar projects that existed before, faced the problem of the speed of synthesis of sound, so to generate a voice, Deep Vice's predecessors required from several minutes to a couple of hours to correctly and naturally reproduce the text in a human voice. The new development of Chinese scientists is based on neural networks and is able to convert text into voice in real time.
Deep Voice can simulate timbre, voice intonation and accent, making them very believable and almost indistinguishable from the real ones, while the voice can be male or female. The developers are sure that their technology can be used as digital assistants, used to record voices in ideograms or even use them for simultaneous translation of films by subtitles.
"This is a real breakthrough from a technical point of view, because we were able to solve the most complicated problem, synthesizing live speech with all its features," says Leo Zoo, one of the authors of the project.
The creators of the algorithm explained that the Deep Voice project is inspired by similar developments, that's just all of its components work under the control of a neural network, while using fairly simple functions in the work, which makes their algorithm very adaptive - the voices can be adjusted "by themselves" accents and other features.
"Deep training has led to a revolution in various fields, such as computer vision and speech recognition, and now is the time and voice synthesis. We are glad that we were able to achieve such results, and we will continue to work to make the text-voice system even more realistic, "Motherboard quotes developers as saying.
Do not forget about our !!
The article is based on materials .
Comments
Post a Comment