?url_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=Procesamiento+de+emociones+por+voz+utilizando+Machine+y+Deep+Learning&rft.creator=Romero+Andr%C3%A9s%2C+Eric&rft.contributor=Arroyo+Montoro%2C+Fernando&rft.contributor=G%C3%B3mez+Canaval%2C+Sandra&rft.subject=Computer+Science&rft.subject=Psychology&rft.description=En+este+proyecto+se+ha+realizado+una+investigaci%C3%B3n+sobre+distintas+t%C3%A9cnicas+de+Machine+Learning+y+Deep+Learning+para+la+construcci%C3%B3n+de+un+clasificador+de+emociones+basado+en+archivos+de+voz.+Tras+indagar+en+los+fundamentos+te%C3%B3ricos+de+las+distintas+t%C3%A9cnicas+se+opt%C3%B3+por+la+construcci%C3%B3n+de+tres+modelos.+Usando+t%C3%A9cnicas+de+Deep+Learning+se+construyeron+dos+redes+neuronales+convolucionales+distintas%2C+mientras+que+con+t%C3%A9cnicas+de+Machine+Learning+se+implement%C3%B3+una+m%C3%A1quina+de+vector+de+soporte.+Para+poder+llevar+a+cabo+el+entrenamiento+de+los+modelos+se+desarroll%C3%B3+un+servicio+de+adquisici%C3%B3n+de+ficheros+de+audio+con+el+que+construir+un+dataset%2C+ya+que+los+datos+necesarios+para+este+prop%C3%B3sito+no+son+muy+abundantes.+Una+vez+construido+el+dataset%2C+se+realiz%C3%B3+un+pre-tratamiento+y+procesamiento+de+los+datos%2C+aplicando+la+trasformada+r%C3%A1pida+de+Fourier+y+la+escala+Mel+para+obtener+los+espectrogramas+de+los+audios.+Estos+espectrogramas+fueron+utilizados+a+continuaci%C3%B3n+para+el+entrenamiento+de+los+modelos+inteligentes.+Utilizando+los+datos+ya+procesados+se+realizaron+m%C3%BAltiples+experimentos+cambiando+las+distribuciones+de+datos+para+entrenamiento+y+testing.+Una+vez+realizados+estos+experimentos+con+los+modelos%2C+se+analiz%C3%B3+el+rendimientos+de+los+mismos+y+se+compararon+los+resultados+obtenidos%2C+llegando+a+la+conclusi%C3%B3n+de+que+todas+las+t%C3%A9cnicas+probadas+podr%C3%ADan+ser+%C3%BAtiles+para+la+resoluci%C3%B3n+del+problema+planteado.++%0D%0A%0D%0AAbstract%3A+%0D%0A%0D%0AIn+this+project%2C+different+Machine+and+Deep+learning+techniques+were+investigated+for+the+construction+of+an+emotion+classiffier+based+on+voice+files.+After+digging+in+the+theoretical+foundation+of+different+techniques%2C+we+opted+for+the+construction+of+three+models.+Using+Deep+Learning+techniques%2C+two+different+convolutional+neural+networks+were+built%2C+while+with+Machine+Learning+techniques%2C+a+support+vector+machine+was+implemented.+In+order+to+train+the+models%2C+an+audio+file+acquisition+service+was+developed+for+building+a+dataset%2C+since+the+data+necessary+for+this+purpose+is+not+very+abundant.+Once+the+dataset+was+built%2C+the+files+were+pre-treated+and+processed+applying+the+fast+Fourier+transform+and+the+Mel+scale+to+obtain+the+spectrograms+of+the+audios.+These+spectrograms+were+then+used+for+the+training+of+the+intelligent+models.+Using+the+data+already+processed%2C+multiple+experiments+were+performed+with+different+data+distributions+for+training+and+testing.+Once+these+experiments+had+been+carried+out%2C+the+performance+of+the+models+was+analyzed+and+compared%2C+concluding+that+all+the+techniques+tested+could+be+useful+for+solving+the+problem.&rft.publisher=E.T.S.I+de+Sistemas+Inform%C3%83%C2%A1ticos+(UPM)&rft.rights=https%3A%2F%2Fcreativecommons.org%2Flicenses%2Fby-nc-nd%2F3.0%2Fes%2F&rft.date=2021-07&rft.type=info%3Aeu-repo%2Fsemantics%2FbachelorThesis&rft.type=Final+Project&rft.type=PeerReviewed&rft.format=application%2Fpdf&rft.language=spa&rft.rights=info%3Aeu-repo%2Fsemantics%2FrestrictedAccess&rft.identifier=https%3A%2F%2Foa.upm.es%2F68234%2F