One of the most attractive functions of music is that it can convey emotion and modulate a listener’s mood. Music can bring to tears, console us when we are grieving and drive us to love. Music information behaviour studies have identified emotion as an important criterion used by people in music searching and organisation. It becomes significant the field of music emotion recognition. Nowadays, is more and more important to retrieve and organise users music, due to the increasing platforms of streaming, which gives the access to a catalog of billions of songs. The automatisation of the recognition of perceived emotion in music allows users to organise and research music in a content-centric fashion. Purpose of this thesis is to find a link between music and emotions during the listening of a song by combining audio and physiological signals analysis. The inclusion of emotions is an hard task, due to the subjective nature of emotion perception. There are problems in the reliability of ground truth data and evaluation of prediction results, which are not troubles in problems as face recognition or speech recognition. One of the most attractive functions of music is that it can convey emotion and modulate a listener’s mood. Music can bring to tears, console us when we are grieving and drive us to love. Music information behaviour studies have identified emotion as an important criterion used by people in music searching and organisation. It becomes significant the field of music emotion recognition. Nowdays, is more and more important to retrieve and organise users music, due to the increasing platforms of streaming, which gives the access to a catalog of billions of songs. The automatisation of the recognition of perceived emotion in music allows users to organise and research music in a content-centric fashion. Purpose of this thesis is to find a link between music and emotions during the listening of a song by combining audio and physiological signals analysis. The inclusion of emotions is an hard task, due to the subjective nature of emotion perception. There are problems in the reliability of ground truth data and evaluation of prediction results, which are not troubles in problems as face recognition or speech recognition.
Una delle funzioni più attrattive della musica è che questa può trasmettere e comunicare emozioni e modulare l’umore di una persona. La musica può provocarci lacrime, consolarci quando siamo tristi, farci innamorare. Gli studi fatti finora sulla musica, affermano che le emozioni sono un criterio importante per la ricerca e l’organizzazione dei brani musicali. Qui diventa fondamentale l’importanza del campo chiamato music emotion recognition. Al giorno d’oggi, diventa sempre più importante il fatto di catalogare e organizzare la musica degli utenti, a causa dell’incremento di piattaforme di streaming musicale, le quali danno accesso ad un numero infinito di brani. L’automatizzazione del riconoscimento delle emozioni percepite in musica, permette all’utente di organizzare e ricercare la musica in una visione più incentrata sul contenuto. Lo scopo di questa tesi è quello di trovare il link tra emozioni percepite durante l’ascolto di un brano musicale attraverso l’analisi del segnale au- dio in primis, ma anche con l’utilizzo di segnali psicologici. L’utilizzo delle emozioni , in generale, è un compito difficile, a causa della natura intrinseca delle emozioni percepite. Ci sono problemi di affidabilità dei dati empirici e la valutazione del modello di predizione, che d’altra parte non sono dei problemi nei casi ben noti di face recognition e speech recognition.
Music emotion detection. A framework based on electrodermal activities
POZZI, GIOELE
2019/2020
Abstract
One of the most attractive functions of music is that it can convey emotion and modulate a listener’s mood. Music can bring to tears, console us when we are grieving and drive us to love. Music information behaviour studies have identified emotion as an important criterion used by people in music searching and organisation. It becomes significant the field of music emotion recognition. Nowadays, is more and more important to retrieve and organise users music, due to the increasing platforms of streaming, which gives the access to a catalog of billions of songs. The automatisation of the recognition of perceived emotion in music allows users to organise and research music in a content-centric fashion. Purpose of this thesis is to find a link between music and emotions during the listening of a song by combining audio and physiological signals analysis. The inclusion of emotions is an hard task, due to the subjective nature of emotion perception. There are problems in the reliability of ground truth data and evaluation of prediction results, which are not troubles in problems as face recognition or speech recognition. One of the most attractive functions of music is that it can convey emotion and modulate a listener’s mood. Music can bring to tears, console us when we are grieving and drive us to love. Music information behaviour studies have identified emotion as an important criterion used by people in music searching and organisation. It becomes significant the field of music emotion recognition. Nowdays, is more and more important to retrieve and organise users music, due to the increasing platforms of streaming, which gives the access to a catalog of billions of songs. The automatisation of the recognition of perceived emotion in music allows users to organise and research music in a content-centric fashion. Purpose of this thesis is to find a link between music and emotions during the listening of a song by combining audio and physiological signals analysis. The inclusion of emotions is an hard task, due to the subjective nature of emotion perception. There are problems in the reliability of ground truth data and evaluation of prediction results, which are not troubles in problems as face recognition or speech recognition.File | Dimensione | Formato | |
---|---|---|---|
Thesis.pdf
accessibile in internet per tutti
Descrizione: Testo della tesi
Dimensione
16 MB
Formato
Adobe PDF
|
16 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/152931