Nofal Tounsi, director of the sales sector at Acapella, a company specializing in providing human performance services for digital voices, said on Saturday that the market for creating digital voices is witnessing rapid development in a way that opens “unlimited horizons” for developers and content makers, but at the same time he warned of “ Unprecedented risks facing society.
Tunisian explained, on the sidelines of the GITEX 2023 technology exhibition in Dubai, that “only 5 minutes of pure audio recorded in a person’s voice is sufficient to create a digital copy of his voice, so that it will be easy to make him pronounce any words in the same language,” noting that “it has become "It is easy to use the voice of any influential or famous person to fabricate facts and spread rumours."
The Belgian company official added, "Caution must be exercised when dealing with generative artificial intelligence techniques," warning that "voice performance professionals have become at risk due to the development of digital voice creation techniques."
Tounsi stated that his company does not intend to enter the market of providing digital voice creation services to individuals, which is largely led by ElevenLabs. He urged companies operating in this market to "adhere to an advanced level of ethical standards, to protect themselves from legal accountability on the one hand, and to prevent the promotion of false information on the other hand."
Tounsi pointed out that “Acapella” is currently working on providing a real-time audio translation service for audio clips between different languages, while maintaining the same voice and performance of the speaker.
Several major companies have entered the voice translation services market, led by Spotify and YouTube.
Acapella digital voices can speak more than 30 languages, including Arabic, English, French, Turkish, Spanish, and Italian, and offer a feature called Voice Smileys, which provides digital voices with the ability to express different emotions, which adds a human character to manufactured voices.
Tounsi said that "Acapella" has worked for 22 years to build and train its own smart models capable of providing human performance of digital voices, to serve the goals of its clients in the marketing and promotional sectors, as well as to support people of determination who lack the ability to speak, to express themselves vocally.
Tounsi pointed out that the company’s work with people of determination is considered the only aspect of its cooperation with individuals, explaining that the company provides the visually impaired and blind with smart voice assistants through its Infovox3 service, which allows the user to interact with any text that appears on his computer screen, saying that “the service reads the text out loud.” Then the user can interact with his device using voice commands.”
Tunisi explained that “Acapella” provides its services to people with speech problems by enabling them to create a digital copy of their voice by recording audio clips with a variety of performances, so that the patient can use that copy to interact with the world naturally later on.