LDV remote voice acquisition and enhancement

Weihong Li, Ming Liu, Zhigang Zhu, Thomas S Huang

Research output: Contribution to journalConference article

Abstract

Laser Doppler vibrometers (LDVs) have been widely used in industry inspection. One of the superior characteristics of an LDV is that it can detect and measure extremely tiny vibration of a target at a large distance, with sensitivity on the order of 1μm/s. On the other hand, we have found that most objects nearby audio sources can be vibrated by the audio waves. These two aspects motivate our research in a new application of the LDVs, namely remote voice detection from surrounding vibrated objects. However, the detected speech signals may be corrupted by many noise sources, such as laser photon noises, target movements, and background acoustic noises (wind, engine sound, etc.). Therefore, speech enhancement algorithms based on Gaussian bandpass and Wiener filters are designed to effectively improve the intelligibility of the noisy voice signals detected by the LDV system Experimental results show that remote voice detection via an LDV is very promising, when choosing appropriate targets close to human subjects and using the proposed enhancement techniques.

Original languageEnglish (US)
Article number1699830
Pages (from-to)262-265
Number of pages4
JournalProceedings - International Conference on Pattern Recognition
Volume4
DOIs
StatePublished - Dec 1 2006
Event18th International Conference on Pattern Recognition, ICPR 2006 - Hong Kong, China
Duration: Aug 20 2006Aug 24 2006

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition

Fingerprint Dive into the research topics of 'LDV remote voice acquisition and enhancement'. Together they form a unique fingerprint.

  • Cite this