| EPSRC Reference: |
GR/M75204/01 |
| Title: |
ROPA: CONTINUOUS-STATE DYNAMICAL SYSTEM MODELS FOR SPEECH RECOGNITION |
| Principal Investigator: |
King, Professor S |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Centre for Speech Technology Research |
| Organisation: |
University of Edinburgh |
| Scheme: |
ROPA |
| Starts: |
01 October 1999 |
Ends: |
31 January 2002 |
Value (£): |
71,143
|
| EPSRC Research Topic Classifications: |
| Human Communication in ICT |
|
|
| EPSRC Industrial Sector Classifications: |
| No relevance to Underpinning Sectors |
|
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
We argue that discrete state Hidden Markov Models are inadequate for modelling observations produced by underlying continuous processes: for example, the movement of the articulators during speech production. We propose to develop a continuous state Markov model. Our recent experience with articulatory measurement data suggests that one approach we can take is to let the (hidden) state represent articulator setting, and constrain the state behaviour by a linear, or possibly non-linear, dynamical system. With hidden observations of trajectories in the state space, training becomes simpler; at recognition time, the state space is hidden. Of course, the articulatory observations could be used only to initialise the model, with the state space being hidden during further training on acoustic-only data.
|
|
Key Findings |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Potential use in non-academic contexts |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Impacts |
|
| Description |
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk |
| Summary |
|
| Date Materialised |
|
|
|
|
Sectors submitted by the Researcher |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
| Project URL: |
|
| Further Information: |
|
| Organisation Website: |
http://www.ed.ac.uk |