A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic Environments
Acoustic event detection and analysis has been widely developed in the last few years for its valuable application in monitoring elderly or dependant people, for surveillance issues, for multimedia retrieval, or even for biodiversity metrics in natural environments. For this purpose, sound source id...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/4/1274 |
| _version_ | 1797411459094806528 |
|---|---|
| author | Daniel Bonet-Solà Rosa Ma Alsina-Pagès |
| author_facet | Daniel Bonet-Solà Rosa Ma Alsina-Pagès |
| author_sort | Daniel Bonet-Solà |
| collection | DOAJ |
| description | Acoustic event detection and analysis has been widely developed in the last few years for its valuable application in monitoring elderly or dependant people, for surveillance issues, for multimedia retrieval, or even for biodiversity metrics in natural environments. For this purpose, sound source identification is a key issue to give a smart technological answer to all the aforementioned applications. Diverse types of sounds and variate environments, together with a number of challenges in terms of application, widen the choice of artificial intelligence algorithm proposal. This paper presents a comparative study on combining several feature extraction algorithms (Mel Frequency Cepstrum Coefficients (MFCC), Gammatone Cepstrum Coefficients (GTCC), and Narrow Band (NB)) with a group of machine learning algorithms (<i>k</i>-Nearest Neighbor (kNN), Neural Networks (NN), and Gaussian Mixture Model (GMM)), tested over five different acoustic environments. This work has the goal of detailing a best practice method and evaluate the reliability of this general-purpose algorithm for all the classes. Preliminary results show that most of the combinations of feature extraction and machine learning present acceptable results in most of the described corpora. Nevertheless, there is a combination that outperforms the others: the use of GTCC together with kNN, and its results are further analyzed for all the corpora. |
| first_indexed | 2024-03-09T04:45:24Z |
| format | Article |
| id | doaj.art-daa945a0cf224668aa79b29870e6652b |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T04:45:24Z |
| publishDate | 2021-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-daa945a0cf224668aa79b29870e6652b2023-12-03T13:16:31ZengMDPI AGSensors1424-82202021-02-01214127410.3390/s21041274A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic EnvironmentsDaniel Bonet-Solà0Rosa Ma Alsina-Pagès1Grup de Recerca en Tecnologies Mèdia (GTM), La Salle—URL, c/Quatre Camins, 30, 08022 Barcelona, SpainGrup de Recerca en Tecnologies Mèdia (GTM), La Salle—URL, c/Quatre Camins, 30, 08022 Barcelona, SpainAcoustic event detection and analysis has been widely developed in the last few years for its valuable application in monitoring elderly or dependant people, for surveillance issues, for multimedia retrieval, or even for biodiversity metrics in natural environments. For this purpose, sound source identification is a key issue to give a smart technological answer to all the aforementioned applications. Diverse types of sounds and variate environments, together with a number of challenges in terms of application, widen the choice of artificial intelligence algorithm proposal. This paper presents a comparative study on combining several feature extraction algorithms (Mel Frequency Cepstrum Coefficients (MFCC), Gammatone Cepstrum Coefficients (GTCC), and Narrow Band (NB)) with a group of machine learning algorithms (<i>k</i>-Nearest Neighbor (kNN), Neural Networks (NN), and Gaussian Mixture Model (GMM)), tested over five different acoustic environments. This work has the goal of detailing a best practice method and evaluate the reliability of this general-purpose algorithm for all the classes. Preliminary results show that most of the combinations of feature extraction and machine learning present acceptable results in most of the described corpora. Nevertheless, there is a combination that outperforms the others: the use of GTCC together with kNN, and its results are further analyzed for all the corpora.https://www.mdpi.com/1424-8220/21/4/1274acoustic sensoracoustic event detectioncorporafeature extractionmachine learning |
| spellingShingle | Daniel Bonet-Solà Rosa Ma Alsina-Pagès A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic Environments Sensors acoustic sensor acoustic event detection corpora feature extraction machine learning |
| title | A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic Environments |
| title_full | A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic Environments |
| title_fullStr | A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic Environments |
| title_full_unstemmed | A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic Environments |
| title_short | A Comparative Survey of Feature Extraction and Machine Learning Methods in Diverse Acoustic Environments |
| title_sort | comparative survey of feature extraction and machine learning methods in diverse acoustic environments |
| topic | acoustic sensor acoustic event detection corpora feature extraction machine learning |
| url | https://www.mdpi.com/1424-8220/21/4/1274 |
| work_keys_str_mv | AT danielbonetsola acomparativesurveyoffeatureextractionandmachinelearningmethodsindiverseacousticenvironments AT rosamaalsinapages acomparativesurveyoffeatureextractionandmachinelearningmethodsindiverseacousticenvironments AT danielbonetsola comparativesurveyoffeatureextractionandmachinelearningmethodsindiverseacousticenvironments AT rosamaalsinapages comparativesurveyoffeatureextractionandmachinelearningmethodsindiverseacousticenvironments |