Georgia Institute of Technology
Multi-Sensor Noise Suppression and Bandwidth Extension for Enhancement of Speech
Abstract
dc:description.abstractSpeech enhancement has been an active research problem for decades and continues to be an important problem. This is made even more true by the proliferation of portable devices having audio input capabilities. In the presence of noise, both the quality and intelligibility of speech signals have been significantly deteriorated. The proposed research are the frameworks for improving the quality/intelligibility of the degraded speech: 1) a single-channel noise suppression system based on perceptual speech detection 2) multi-sensor noise suppression system for acoustic harsh environments based on non-air conductive sensors 3) a speech bandwidth extension system for telephone speech Significant improvement in both speech intelligibility and quality from the proposed frameworks are indicated from extensive experiments, inlcuding MOS, DRT, speech recognition task, and log spectral distortion.
Degree
thesis:*- Department dc:contributor.department
- Electrical and Computer Engineering
- Grantor dc:publisher
- Georgia Institute of Technology
- Year dc:date.issued
- 2006
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Hu, Rongqiang
- Advisor dc:contributor.advisor
-
- Anderson, David V.
- Committee members dc:contributor.committeemember
-
- Chin-Hui Lee
- Gordon Stuber
- Mark Clements
- Xu-Yan Chen
Subjects
dc:subject × 5Rights
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- http://hdl.handle.net/1853/10452