Visible Speech Perception Research Page

Christopher S. Campbell

Christopher Campbell is a graduate student in the Program in Experimental Psychology at the University of California - Santa Cruz. His research interests include visible speech perception (lipreading or speechreading), auditory speech perception, speech production, and computational models. He received a BA in Psychology (1993), a BA in Philosophy (1993) from Purdue University and an MS in Experimental Psychology from the University of California. For more info go here.
 
Christopher Campbell 

Professional Research 
 
 
 
Resources 

News
 

 
Labs, People, Journals &
Organizations
 
Links to More Resources
 
 
 
 
 
 
 
 
 
 

Current Research

  Current!! 
 

  • Visible Speech Perception: Influence of Spatial Quantization.


Believe it or not, everyone is influenced by the movements of a speaker's face while in a face to face conversation. Thats right, even people who have never been trained to lipread use information from the speaker's face to perceive speech. In Psychology we call the influence of visible speech on speech perception the McGurk effect. The goal of the present experiment was to determine exactly what visual information in a synthetic talking head everyday people use to perceive speech. To do this, we elminated channels or frequencies of image data using the technique of spatial quantization and measured the effect on performance. The degree that the proportion of correct syllable identification was changed by the data removed informs us as to the information content of that data for visible speech perception. Another goal of the present experiment was to mathematically model the perceptual confusions of subjects using a set of apriori features which are integrated according to two different theories. The fuzzy logical model of perception (FLMP) predicts that features are integrated multiplicatively while the additive model of perception (AMP) predicts an additive integration.



    What are some of the practical aspects of this research?

  1. Improvement of programs that teach the deaf to speechread (lipread).
  2. Higher image compression of natural and synthetic talking heads.
  3. Speech recognition enhancement through the inclusion of visual data with auditory data.
  4. Increased intelligibility of computer generated talking heads.

The above picture shows the five levels of spatial quantization used during the experiment. The top panels demonstrate a /va/ articulation while the bottom panels show a /wa/ articulation.

For more information on speechreading, go here.

Publications

  • Campbell, C. S., Massaro, D. W., & Schwarzer, G. (in press). Face perception: An information processing framework. In M. J. Wenger and J. Townsend (Eds.), Computational, geometric, and process perspectives on facial cognition: Contexts and Challenges. MIT Press.
     
  • Campbell, C. S., Massaro, D. W. (1998). Robust visible speech perception and automatic face processing. In H. Wechsler, J. Phillips, V. Bruce, F. Fogelman Soulii, T. Huang (Eds.), Face recognition: From theory to applications. Berlin: Springer-Verlag.
     
  • Campbell, C. S., Massaro, D. W. (1997). Perception of visible speech: Influence of spatial quantization. Perception, 26, 627-644.

  Conference papers and posters
     
  • Campbell, C. S., Massaro, D. W. (1997). Relational/Featural and global/local information dimensions in visible speech perception. The Fifth Annual Workshop on Object Perception and Memory. Philadelphia, PA. (November, 1997).
     
  • Massaro, D. W., Campbell, C. S., Berger, M. A., & Cohen, M. M. (1996). Speechreading talking faces. The 3rd Joint Meeting of the Acoustical Society of America-Acoustical Society of Japan. Honolulu, Hawaii. (December, 1996).
     
  • Campbell, C. S. & Massaro, D. W. (1997). Robustness of visible speech perception. The NATO ASI Conference on Face Recognition. Stirling, Scotland. (June, 1997).
     
__________
 
Up to Perceptual Science Lab MainPage Back to Main Page
 

Comments and suggestions are greatly encouraged:
E-mail: ccampbel@fuzzy.ucsc.edu

Please send corrspondence and requests for reprints to:

Christopher S. Campbell
Psychology Department
UC-Santa Cruz
Santa Cruz CA, 95064
This site is maintained by the UCSC Perceptual Science Laboratory. Send comments and questions to Christopher Campbell ccampbel@cats.ucsc.edu or Michael Cohen mmcohen@ranx.ucsc.edu

Last Updated: 19 Sepember 1998
perception perception perception perception perception perception perception perception perception perception perception perception speech speech speech viseme viseme viseme phoneme phoneme speechreading speechreading speechreading lipreading lipreading lipreading visual visual visual modelling modelling modelling compression compression compression information information psychophysics psychophysics psychophysics psychophysics psychophysics psychophysics psychophysics psychophysics psychophysics psychophysics psychophysics psychophysics