Created by a speech synthesizer based on AI that can play any voice

Almost all currently existing virtual assistants, no matter how advanced features they may possess, have very strange “machine” voice, that with nothing to confuse. This occurs from the fact that the voice synthesizers used for speech set of pre-recorded phrases, sounds and combinations thereof. As suggested by some experts in this area, improving sound quality can be achieved by using artificial intelligence, and demonstrated the company’s Lyrebird from Montreal. Their development can reproduce the voice of any person.

To follow the voice system takes only a few seconds of audio recording of voice required of a person on the basis of which will create an audio snippet. Accurate imitation of the voice is possible through the use of neural networks based on artificial intelligence and operating according to the same principles as the neural networks of the human brain. The AI learns to recognize features of human speech, and then these data are already used for the synthesis of an artificial voice. Now the new system is still not without its flaws: there are problems with the intelligibility of spoken speech, there are “voice artifacts” and some other signs pointing that a statement made machine. However, they can be easily eliminated in the future, because now the system works in real-time. According to one of the sponsors Jose Sotelo,

“Our program was trained on a large number of audio fragments of the speeches of thousands of different people. The information obtained is compressed to form a kind of “voice DNA”, which is a digital key. Then, on the basis of this key the system can reproduce any word, even those that were not involved in the process of learning.”

The authors of the draft understand that the proper level of development of this technology to avoid security issues. For example, to bypass systems identify the user by voice. Company representatives Lyrebird comparing his invention with the invention of photoshop. After you create a software package from Adobe, it became difficult to trust the images on the screen. Now you can’t trust the voice.

“We understand that due to the high level of development of modern technologies such a voice synthesizer come to light sooner or later. We encourage everyone to begin to refuse to accept as evidence a variety of audio recordings, as well as using remedies on the basis of voice.”

In any case, there’s nothing to worry much because the system is very crude, and synthetic voices still sound “robotic notes”. To see how the system Lyrebird reproduces the voices of Barack Obama, Donald trump and several other political figures, you can follow this link.

Created by a speech synthesizer based on AI that can play any voice
Vladimir Kuznetsov


Date:

by