Currently available projects

« Back

Gesture Generation and Recognition: Sign Language Analysis and Synthesis

Information

  • Start date: October 2013
  • Programme: PhD
  • Mode of Study: Full Time
  • Studentship Length: 3 years

How to Apply

Fees & Funding

  • Funding Status: Competition Funded Project (EU Students Only)
    Further Details
  • Funding Source: Funding is available from a number of different sources
  • Funding Conditions:

    Funding is available to EU students. If funding is awarded for this project it will cover tuition fees and stipend for UK students. EU students may be eligible for full funding, or tuition fees only, depending on the funding source.

  • Fees: Fees Information (Opens in new window)

Entry Requirements

  • Acceptable First Degree:

    Computer Science (Graphics, Vision processing), Mathematics, Linguistics

  • Minimum Entry Standard: 2:1

Project Description

The Virtual Humans Group at UEA has an international reputation in Avatar Animation with specific application to Deaf Sign Language. The main focus of the work is the development of notations for describing signing gestures and generation of realistic animation from transcriptions using such notations.

Through the Dicta-Sign project, the group has access to a developing parallel corpus of sign language materials for four European sign languages, annotated at varying levels of detail with information that can be used both for linguistic analysis and resynthesis of the original performance. A number of tools have been developed using computational image and linguistic processing to automate a range of annotation tasks.

There are opportunities for PhD study for students with a strong background in one or more of computer vision processing, computer graphics, computing science, and computational linguistics. The focus of the work will be to survey the limitations of existing animation systems and research additional capabilities required for an improved gesture notation. The aim will be to enhance the quality and realism of generated sign language, based on analysis of corpus data through vision and linguistic processing.

Potential applicants interested in this topic or related research are encouraged to contact the primary supervisor and see http://vh.cmp.uea.ac.uk  for further information

References

"Towards the Integration of Synthetic SL Animation with Avatars into Corpus Annotation Tools"

Ralph Elliott, Javier Bueno, Richard Kennaway, and John Glauert
in Proceedings Workshop on Corpora and Sign Language Technologies (CSLT), LREC 2010, Malta, pp84-87, May 2010.http://www.sign-lang.uni-hamburg.de/lrec2010/lrec_cslt_01.pdf

"Linguistic Modelling and Language-Processing Technologies for Avatar-based Sign Language Presentation". Elliott, R., Glauert, J.R.W., Kennaway, J.R., Marshall, I. and Safar, E.
UAIS (Universal Access in the Information Society) Journal, Springer, Vol. 6, No. 4, pp375-391, February 2008.

"Providing Signed Content on the Internet by Synthesized Animation"
Kennaway, R., Glauert, J.R.W., and Zwitserlood, I.  Transactions on Computer-Human Interaction, ACM Press, Vol. 14, No. 3, 29pp, September 2007.

Publications and Public Project Deliverables of the Dicta-Sign project.  http://dictasign.eu



Apply online