Saliency-based modeling of acoustic scenes using sparse non-negative matrix factorization

Abstract : The modelling of auditory scenes is a challenging task in Computational Auditory Scene Analysis. A method based on sparse Non-negative Matrix Factorization that can be used with no prior knowledge of the audio content to establish the similarity between scenes is proposed. The method is evaluated on a corpus of soundscapes of train stations issued from a perceptual study and results are compared with the human perception. The proposed method, by being able to focus on salient events within the scene, achieves better performances than a state-of-the-art Bag-of-Frames approach though not reaching the human performances.
Document type :
Conference papers
Complete list of metadatas

Cited literature [9 references]  Display  Hide  Download
Contributor : Arshia Cont <>
Submitted on : Friday, January 31, 2014 - 12:07:08 PM
Last modification on : Saturday, October 12, 2019 - 1:20:04 PM
Long-term archiving on : Sunday, April 9, 2017 - 4:18:32 AM


Publisher files allowed on an open archive


  • HAL Id : hal-00940075, version 1


Benjamin Cauchi, Mathieu Lagrange, Nicolas Misdariis, Arshia Cont. Saliency-based modeling of acoustic scenes using sparse non-negative matrix factorization. Workshop on Image and Audio Analysis for Multimedia Interactive, Jul 2013, Paris, France. ⟨hal-00940075⟩



Record views


Files downloads