Lightweight Annotation and Class Weight Training for Automatic Estimation of Alarm Audibility in Noise - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Lightweight Annotation and Class Weight Training for Automatic Estimation of Alarm Audibility in Noise

Résumé

In an effort to improve occupational health and safety, we recently proposed an approach to assess the audibility of acoustic danger signals. It is based on the use of a binary classifier trained on perceptual data to predict the audibility of acoustic alarms in audio clips. In the present article, we first investigate the impact of label noise in the training data induced by a flexible annotation procedure on the model performance. We show that a lighter annotation procedure at training still allows for reaching close to human performance at test time. Besides, threshold selection is a crucial aspect in our application as it can have a direct impact on user safety. We thus explore class weight to train a model that allows for a more robust decision threshold selection, ensuring a low false positive rate.
Fichier principal
Vignette du fichier
F.Effa - Lightweight Annotation and Class Weight Training for Automatic Estimation of Alarm Audibility in Noise.pdf (239.46 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04385004 , version 1 (10-01-2024)

Licence

Paternité

Identifiants

Citer

François Effa, Romain Serizel, Jean-Pierre Arz, Nicolas Grimault. Lightweight Annotation and Class Weight Training for Automatic Estimation of Alarm Audibility in Noise. ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing, Jun 2023, Rhodes Island, Greece. pp.1-5, ⟨10.1109/ICASSP49357.2023.10094730⟩. ⟨hal-04385004⟩
23 Consultations
11 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More