Water flow detection from a wearable device with a new feature, the spectral cover - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2012

Water flow detection from a wearable device with a new feature, the spectral cover

Résumé

This paper presents a new system for water flow detection on real life recordings, and its application to medical context. The recognition system is based on an original feature for sound event detection in real life. This feature, called "spectral cover" shows an interesting behaviour to recognize water flow in a noisy environment. An experiment is realized with more than 7 hours of videos recorded by a wearable device. The system based on thresholds obtains good results for the water flow event recognition (Fmeasure of 66%). A comparison with classical approaches using MFCC or low levels descriptors and GMM classifiers is done to attest the good performance of our system. Adding the spectral cover to low levels descriptors also improve their performance and confirms the interest to this feature.
Fichier principal
Vignette du fichier
2012_CBMI_Patrice.pdf (303.32 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03684272 , version 1 (01-06-2022)

Identifiants

Citer

Patrice Guyot, Julien Pinquier, Régine André-Obrecht. Water flow detection from a wearable device with a new feature, the spectral cover. 10th International Workshop on Content-Based Multimedia Indexing (CBMI 2012), Jun 2012, Annecy, France. pp.1-6, ⟨10.1109/CBMI.2012.6269814⟩. ⟨hal-03684272⟩
33 Consultations
31 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More