?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=Sastre+Gallardo%2C+Alberto&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+investigaron+diferentes+t%C3%A9cnicas+de+Machine+Learning+y+Deep+Learning+para+la+construcci%C3%B3n+de+un+clasificador+de+emociones+en+base+a+archivos+de+audio.+Se+explica+el+funcionamiento+te%C3%B3rico+de+los+clasificadores+de+emociones+construidos.+Para+Deep+Learning+se+construy%C3%B3+un+red+neuronal+convolucional+mientras+que+para+Machine+Learning+se+implement%C3%B3+una+m%C3%A1quina+de+vector+de+soporte.%0D%0APara+poder+realizar+el+entrenamiento+de+ambos+modelos+se+desarroll%C3%B3+un+servicio+de+adquisici%C3%B3n+de+ficheros+de+audio+con+el+objetivo+de+construir+un+dataset+propio%2C+ya+que%0D%0Alos+datos+en+este+formato+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+se+obtuvo+el++espectrograma+de+los+audios%2C+que+finalmente+fueron+utilizados+para+el+entrenamiento+de+los+modelos+inteligentes.+Una+vez+realizado+el+entrenamiento+de+ambos+modelos+se+analizaron+sus+rendimientos+en+relaci%C3%B3n+a+un+mismo+conjunto+de+datos+y+se+compararon+sus+resultados%2C+llegando+a+la+conclusi%C3%B3n+de+que+ambas+metodolog%C3%ADas+podr%C3%A1n+ser+%C3%BAtiles+para+la+resoluci%C3%B3n+del+problema.%0D%0AAbstract%3A%0D%0AIn+this+project%2C+different+Machine+Learning+and+Deep+Learning+techniques+for+the+construction+of+an+emotion+classifier+based+on+audio+files+were+investigated.+The+theoretical+operation+of+the+constructed+emotion+classifiers+is+explained.+For+Deep+Learning+a+convolutional+neural+network+was+built+while+for+Machine+Learning+a+support+vector+machine+was+implemented.+In+order+to+train+both+models%2C+an+audio+file+acquisition+service+was+developed+with+the+aim+of+building+an+own+dataset%2C+since+the+data+in+this+format+is+not+very+abundant.+Once+the+dataset+was+built%2C+a+pre-treatment+and+data+processing+was+performed%2C+applying+the+fast+Fourier+transform+and+the+Mel+scale%2C+the+audio+spectrogram+was+obtained%2C+which+was+finally+used+for+the+training+of+intelligent+models.+Once+the+training+of+both+models+was+carried+out%2C+their+performances+in+relation+to+the+same+dataset+were+analyzed+and+their+results+were+compared%2C+concluding+that+both+methodologies+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=2020-07&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%2F64331%2F