Skip to Main content Skip to Navigation
Conference papers

SOUND DETECTION AND CLASSIFICATION THROUGH TRANSIENT MODELS USING WAVELET COEFFICIENT TREES

Michel Vacher 1, * Dan Istrate 1 Jean-François Serignat 1
* Corresponding author
1 CLIPS-IMAG - Equipe GEOD, Groupe d'étude sur l'oral et le dialogue
LIG - Laboratoire d'Informatique de Grenoble
Abstract : Medical Telesurvey needs human operator assistance by smart information systems. Usual sound classification may be applied to medical monitoring by use of microphones in patient's habitation. Detection is the first step of our sound analysis system and is necessary to extract the significant sounds before initiating the classification step. This paper proposes a detection method using transient models, based upon dyadic trees of wavelet coefficients to insure short de-tection delay. The classification stage uses a Gaussian Mix-ture Model classifier with classical acoustical parameters like MFCC. Detection and classification stages are evalu-ated in experimental recorded noise condition which is non-stationary and more aggressive than simulated white noise and fits with our application. Wavelet filtering methods are proposed to enhance performances in low signal to noise ra-tios.
Document type :
Conference papers
Complete list of metadatas

Cited literature [10 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01088253
Contributor : Michel Vacher <>
Submitted on : Thursday, November 27, 2014 - 5:05:28 PM
Last modification on : Thursday, November 19, 2020 - 12:59:57 PM
Long-term archiving on: : Friday, April 14, 2017 - 10:02:02 PM

File

2004_EUSIPCO_Vacher.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01088253, version 1

Collections

CNRS | LIG | UGA

Citation

Michel Vacher, Dan Istrate, Jean-François Serignat. SOUND DETECTION AND CLASSIFICATION THROUGH TRANSIENT MODELS USING WAVELET COEFFICIENT TREES. 12th European Signal Processing Conference (EUSIPCO), Sep 2004, Vienne, Austria. pp.1171-1174. ⟨hal-01088253⟩

Share

Metrics

Record views

83

Files downloads

225