@phdthesis{Schorradt2025, author = {Schorradt, Martin}, title = {The semantic and spatio-temporal structure of audible speech}, doi = {10.26127/BTUOpen-7087}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus4-70875}, school = {BTU Cottbus - Senftenberg}, year = {2025}, abstract = {The goal of the present thesis is to assess whether it is possible to rejuvenate voices of adults to a certain desired age. Therefore, in a set of recordings, two of the main voice features (i.e., pitch and formants) were manipulated in a systematic way. Since children do not only sound different from adults but also articulate in an other manner the resynthesized recordings were made unintelligible. The present thesis presents an evaluation of different entrenched techniques for semantic removal and furthermore, aimes to offer a new approach based on a morphological operator. To not just give insight about the capability of the conducted manipulations but also to provide valuable results in the field of perception of speech, a series of perceptual experiments was conducted.}, subject = {Age synthesis; Age regression; Speech synthesis; Speech processing; Semantic space; Alterssynthese; Altersregression; Sprachsynthese; Sprachverarbeitung; Semantischer Raum; Sprachverarbeitung; Automatische Sprachproduktion; Stimme; Lebensalter; Semantischer Raum; Wahrnehmung; Experiment}, language = {en} }