Voice technology has come a long way, but with the implementation of artificial intelligence techniques, it has made significant advances.
It currently comprises three main processes: speech recognition, voice synthesis, and voice cloning. The first processes people’s speech and automatically identifies what they want to say with a high degree of understanding, a technique implemented in personal assistants like Siri and Alexa, home devices related to these assistants, translation services, and more.
The second, no less important, is voice synthesis, with its ability to convert text into human voices, used in transcription applications, translation services, and the generation of artificial voices.
Finally, there is voice cloning, a field of voice technology that is less commonly used than the previous ones but has great potential in the entertainment sector, as it allows for the creation of identical copies of the voice of a living or deceased person. This serves as a framework for creating new versions of songs, generating voices for video games, among other things.
However, it is also a source of controversy, as it can be used for fraudulent operations such as fake voice calls or messages.
How to clone your voice or someone else’s voice with AI
The good news is that artificial intelligence voice cloning tools can be run directly from your web browser. One of them, and perhaps the one that offers the best user experience, is the platform PlayHT: https://play.ht/, which includes a voice cloning tool among its AI-powered text-to-voice generators.
To generate a clone of your own voice or someone else’s voice, simply upload a sample of the recorded voice to their servers in an audio file, preferably captured with high-quality microphones. To do this, just go to their Voice Cloning section and click the Clone a voice now button.
As is customary on web platforms, you must register as a new user before using their services. After this, they provide you with 2500 words to experiment with how they sound with your cloned voice. If you want more words, you will need to subscribe to a subscription plan.
Image created with Luzia from Telegram