CereProc in collaboration with Accenture has created the world’s first non-binary synthetic voice
CereProc Ltd, collaborating with Accenture Labs has created the world’s first non-binary synthetic voice that combined male and female data from transgender and non-binary individuals for the global digital assistant market. The companies have released all materials to generate the voice to the Open Source community.
As soon as you give an artificial system a voice you give it a character. The majority of voice assistants such as Siri, Cortana and Alexa sound female. If you ask any voice assistants what gender they are, or if they’re female, they’ll all respond that they don’t have a gender. Gender is a design choice and often an implicit one. This design choice can reinforce negative gender stereotypes. However, an artificial system running on a computer doesn’t have a biological gender. Dr Matthew Aylett, Chief Scientific Officer states: “At CereProc, we want software designers and engineers to choose the character they ascribe to a system by designing the voice they want to use. As part of ground-breaking work, we have carried out with Accenture, we now add perceived gender to that set of design choices”. The release of the world’s first non-binary synthetic voice not only gives non-binary people a voice that represents them, but it sparks the conversation on social issues and could mark the end of gender bias in voice assistants.
By applying sophisticated machine learning modelling techniques with CereProc’s Neural TTS system, CereWave A.I., CereProc was able to mix genders and non-binary prosody across voices, creating ‘Sam’. Producing a complex gender voice is not just a matter of changing the pitch of the voice but of modelling speech styles and the vocal tract. The technology developed in this project has given CereProc the ability to mix genders across voices and even take a voice changing its gender completely.
Below is an example of our non-binary synthesis voice
Sam Example 1:
“While gender-neutral voice samples have been released previously, Sam is the first non-binary AI-based digital voice solution that can be embedded into any software solution to speak text in a human-sounding voice,” said Marc Carrel-Billiard, senior managing director and Technology Innovation lead at Accenture. “This work is a great example of how technology and human creativity can come together and spark a moment of societal change that can benefit many people. We believe it’s essential that voice assistants more accurately represent the diversity of our global population.”
Paul Welham, Chief Executive Officer of CereProc concluded:
“By creating a non-binary voice with Accenture, CereProc are challenging strongly held mainstream views in IT that a synthetic voice has to be clearly male or female. One of CereProc’s key aims since inception has been to empower application designers to disrupt the status quo in speech, with a non-binary voice we wish to raise awareness of this important issue in the new wave of A.I. based digital voice systems being developed with our partners.”