CereProc is collaborating in a research project and touring exhibition that explores the principles for conversational emotion and challenges involved in artificial voices.
A team from speech technology company CereProc, collaborating with fellow scientists and speech practitioners as part of the global interdisciplinary Creative Speech and Technology (CreST) network, is set to participate in a research study conceived by the network's Voice Expressivity and Emotion Group (VEEG) to explore challenges and new practices involved in synthetic voice production and emotional conversation.
Taking up their position alongside industry specialists from around the world, CereProc's experts will engage in the wider project while focusing work on the design and build of the central experimental chat bot-named the Conversational Kiosk installation-before showcasing the research as part of a touring exhibition that aims to heighten public engagement with computer speech research and encourage uptake of speech technology in a number of applications.
Led by Dr Alistair Edwards from the University of York and Dr Christopher Newell from the University of Hull, the project named Articulate: The Art and Science of Synthetic Speech - supported by Engineering and Physical Sciences Research Council (EPSRC) and the Arts Council England (ACE) - explores the nature of emotional conversation using the interactive, chat bot powered by CereProc's text-to-speech technology, and features exhibits created by scientists and artists using state-of-the-art speech technologies.
Alongside text-to-speech technology and synthetic speech expertise delivered by CereProc, the CreST network blended multiple disciplines involving specialists from augmentative and alternative communication (AAC) for disabled users, public voice-announcements and telephone self-service systems, musical composition, human-computer interaction (HCI), and artificial intelligence (AI) to reach its research objectives.
"Speech-synthesis technology allows those who have lost their voice through illness or disability to communicate verbally," Edwards said in a statement. "However, there is an urgent need for this type of technology to be more widely available and for it to be more reliable and personal."
Chris Pidcock, chief voice engineer at CereProc, explained: "We're very proud to be involved in the CreST network, and to have our emotional text-to-speech voices applied to such an innovative and valuable project. Being given the opportunity to collaborate so closely with such a dedicated group of creative individuals has been and continues to be an absolute privilege.
"With the tour kicking off on Monday December 3, we're looking forward to exhibiting the interactive chatbot, engaging with a wide variety of audiences across the country in the coming months, and ultimately raise awareness of synthetic speech and its wide range of capabilities for the future," he continued.
Edwards added: "By providing more public engagement with computer speech research using the creative and performing arts, CreST aims to help people gain more understanding of the benefits and limitations of this type of technology."
Commenced in March 2011 as part of a two-year collaborative partnership between members of the CreST network, the forthcoming touring exhibition is a demonstration of the culmination of works developed within the network and will offer special activities for the public supported by an interactive Web site.
The CreST VEEG group exhibition will visit various locations in York in the U.S. and Sheffield and Hull in the U.K. on December 3, 4 and 5 respectively, before being given a temporary residence at the Woodend Gallery, The Crescent in Scarborough, Englasnd, between January 22 and 26.