In this work, we propose a machine learning-based classification approach aiming at identifying real-world sounds recorded in construction sites. The proposed approach is based on the leaky version of the Echo State Network (ESN), and it has been tested on a real-world dataset composed of recordings of five vehicles and tools usually used in construction sites. The implemented leaky version of the ESN exploits different spectral features as input. After the description of the proposed approach, we provide some numerical results obtained on the recorded signals. The overall accuracy on the test set, after the integration of a majority voting approach, is up to 95.3%, comparable to other state-of-the-art machine learning methods demonstrating the effectiveness of the approach.

Leaky Echo State Network for Audio Classification in Construction Sites / Scarpiniti, Michele; Bini, Edoardo; Ferraro, Marco; Giannetti, Alessandro; Comminiello, Danilo; Lee, Yong-Cheol; Uncini, Aurelio. - (2023), pp. 183-193. - SMART INNOVATION, SYSTEMS AND TECHNOLOGIES. [10.1007/978-981-99-3592-5_18].

Leaky Echo State Network for Audio Classification in Construction Sites

Michele Scarpiniti;Edoardo Bini;Marco Ferraro;Alessandro Giannetti;Danilo Comminiello;Aurelio Uncini
2023

Abstract

In this work, we propose a machine learning-based classification approach aiming at identifying real-world sounds recorded in construction sites. The proposed approach is based on the leaky version of the Echo State Network (ESN), and it has been tested on a real-world dataset composed of recordings of five vehicles and tools usually used in construction sites. The implemented leaky version of the ESN exploits different spectral features as input. After the description of the proposed approach, we provide some numerical results obtained on the recorded signals. The overall accuracy on the test set, after the integration of a majority voting approach, is up to 95.3%, comparable to other state-of-the-art machine learning methods demonstrating the effectiveness of the approach.
2023
Smart Innovation, Systems and Technologies
978-981-99-3591-8
978-981-99-3592-5
audio processing; construction sites; environmental sound classification; leaky echo state network (ESN); machine learning
02 Pubblicazione su volume::02a Capitolo o Articolo
Leaky Echo State Network for Audio Classification in Construction Sites / Scarpiniti, Michele; Bini, Edoardo; Ferraro, Marco; Giannetti, Alessandro; Comminiello, Danilo; Lee, Yong-Cheol; Uncini, Aurelio. - (2023), pp. 183-193. - SMART INNOVATION, SYSTEMS AND TECHNOLOGIES. [10.1007/978-981-99-3592-5_18].
File allegati a questo prodotto
File Dimensione Formato  
Scarpinti_postprint_Leaky2023.pdf

embargo fino al 01/09/2025

Note: postprint
Tipologia: Documento in Post-print (versione successiva alla peer review e accettata per la pubblicazione)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 337.12 kB
Formato Adobe PDF
337.12 kB Adobe PDF   Contatta l'autore
Scarpiniti_Leaky_2023.pdf

solo gestori archivio

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 185.29 kB
Formato Adobe PDF
185.29 kB Adobe PDF   Contatta l'autore

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11573/1693486
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact