| EPSRC Reference: |
GR/G64916/01 |
| Title: |
NOISE CANCELLATION FOR IN VEHICLE SPEECH RECOGNITION (SPEECH IN VEHICLES) |
| Principal Investigator: |
Elliott, Professor S |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Inst of Sound and Vibration Research |
| Organisation: |
University of Southampton |
| Scheme: |
Standard Research (Pre-FEC) |
| Starts: |
09 June 1992 |
Ends: |
08 October 1995 |
Value (£): |
92,383
|
| EPSRC Research Topic Classifications: |
| Human Communication in ICT |
|
|
| EPSRC Industrial Sector Classifications: |
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
To investigate microphone arrangements and signal processing techniques for clear speech reception in noisy vehicle interiors, with the aim of improving the performance of speech recognisers used in cars.(This work has been carried out in cooperation with Lotus Engineering and Marconi Speech and Information Systems, as part of JFIT project ITD3/1/3062).Progress:Initially, the use of single microphones for in-car speech reception was studied. Headset microphones were rejected as too inconvenient for the user. The best type of microphone and its positioning on the car interior were determined by measurements of speech-signal-to-noise-ratio (SSNR), using recordings of speech made in cars under various driving conditions. Several single-channel speech enhancement techniques were tested, using the in-car speech recordings as input data.The use of microphone arrays was studied next. The number of microphones and their positioning were considered, and several types of processing of their outputs were conceived and tested. Again, in-car speech recordings were used to measure the performances of the various array configurations. Performances were judged from both SSNRs and speech recognition error rates. This work has shown that considerable improvements in the performance of in-vehicle speech recognisers can be achieved, by correct design of the speech microphone(s), and by noise-reducing processing of the microphone output(s). The most promising processing techniques are based on adaptive filtering, using error criteria matched to the speech recognition algorithm. Development of these techniques is continuing.
|
|
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.soton.ac.uk |