Legacy: Talking Heads

This legacy webpage from 1998 was originally developed to provide an overview of some of the early international efforts to create talking heads (cognitive and computational models of audio-visual speech production), the historical antecedents of this effort, and related work. It was developed by Philip Rubin of Haskins Laboratories in New Haven, Connecticut, and Eric Vatikiotis-Bateson, then of ATR Research Laboratories, Kyoto, Japan and later of the Linguistics Department of the University of British Columbia. The original inspiration for this webpage was the AVSP'97 meeting (the European Tutorial and Workshop on Audio-Visual Speech Processing) held in Rhodes (Greece), 26-27 September 1997. This was a satellite meeting of the 5th ESCA (now ISCA)-organized EuroSpeech '97 Conference held in Rhodes on September 22-25, 1997. Information is also available for the other AVSP and ISCA meetings.

These meetings led to the creation of AVISA, the Auditory-Visual Speech Association, by Christian Benoît, Denis Burnham, Philip Rubin, and Eric-Vatikiotis Bateson. AVISA (the Auditory-VIsual Speech Association), which was the second ISCA Special Interest Group (SIG) to form, started its official business on December 5, 1998, at the AVSP'98 meeting in Terrigal, New South Wales, Australia. Rubin briefly served as the found President of AVISA, which is now part of ISCA. AVISA was first conceived by Christian Benoît, who wanted to see the auditory-visual speech community come together in a more formal way. Before his untimely death at the age of 42, Christian drafted a proposal to create AVISA, which is dedicated to his memory.

Over the years, the Talking Heads project turned into a placeholder for a book that Rubin and Bateson were working on called, surprisingly, Talking Heads. Unfortunately, Eric Vatikiotis-Bateson died on May 31, 2017, and the project came to a screeching halt. Prior to his death, drafts of the book sections such as:

  • Being Biological;

  • Simulacra: The Early History of Talking Machines;

  • Virtual Vocal Tracts;

  • Articulators - lips, tongue, and jaw;

  • Measuring and Modeling Speech Production;

  • The McGurk Effect: hearing lips and seeing voices;

  • Speechreading;

  • Facial Animation;

  • Avatars;

  • along with Background Information and a Bibliography.

Over the years, much of this material has been covered well by others. One day, Rubin hopes to the web version of this project, including featuring some of Bateson’s fascinating research during his final years. But the book will never appear.

This website was originally dedicated with great love and appreciation to the memory of our friend, Christian Benoît. And now, Eric Vaikiotis Bateson joins the list. Stay tuned to see if I last long enough to add anything.

- - - - Philip Rubin