Steering latent voice representations for voice reconstruction based on a witness’s memory trace

StatusVoR
Simple item page
dc.abstract.enThis article addresses the significant scarcity of forensic tools for generating voice samples for identification parades by proposing a novel method for voice reconstruction without a reference recording. The presented solution serves as an auditory analogue to facial composites, utilizing a hybrid architecture based on machine learning and speech synthesis (XTTS). The methodology combines an iterative algorithm for selecting a base voice candidate with a specialized neural network module that allows for the modification of interpretable acoustic parameters via latent space manipulation. Technical validation confirmed the system’s effectiveness in navigating the voice space and accurately translating physical parameters into vector representations. By shifting the identification burden from error-prone verbal descriptions to direct auditory perception, the system minimizes the verbal overshadowing effect. Consequently, the proposed prototype offers promising practical implications for forensic science, providing a technological foundation for law enforcement to conduct accurate voice lineups based solely on witness memory.
dc.affiliationWYdział Psychologii w Katowicach
dc.affiliationWydział Psychologii w Katowicach
dc.affiliationInstytut Psychologii
dc.contributor.authorOlawski, Jakub
dc.contributor.authorKuś, Filip
dc.contributor.authorDukała, Karolina
dc.contributor.authorWitkowski, Marcin
dc.date.access2026-09-14
dc.date.accessioned2026-09-18T08:04:32Z
dc.date.available2026-09-18T08:04:32Z
dc.date.created2026-06-16
dc.date.issued2026
dc.description.accesstimeat_publication
dc.description.physical45–63
dc.description.sdgPeaceJusticeAndStrongInstitutions
dc.description.versionfinal_published
dc.description.volume145
dc.identifier.doi10.4467/12307483PFS.26.003.23948
dc.identifier.eissn2720-5983
dc.identifier.issn1230-7483
dc.identifier.urihttps://share.swps.edu.pl/handle/swps/2528
dc.identifier.weblinkhttps://ejournals.eu/czasopismo/problems-of-forensic-sciences/artykul/steering-latent-voice-representations-for-voice-reconstruction-based-on-a-witnesss-memory-trace
dc.languageen
dc.languagepl
dc.language.otherpl
dc.pbn.affiliationpsychologia
dc.rightsCC-BY-NC-ND
dc.rights.questionYes_rights
dc.share.articleOPEN_JOURNAL
dc.subject.enforensic acoustics
dc.subject.envoice reconstruction
dc.subject.enmachine learning
dc.subject.envoice lineup
dc.subject.enauditory memory
dc.subject.enlatent space
dc.swps.sciencecloudsend
dc.titleSteering latent voice representations for voice reconstruction based on a witness’s memory trace
dc.title.alternativeKoncepcja metody modyfikacji ukrytych reprezentacji głosu w celu jego rekonstrukcji na podstawie śladu pamięciowego świadka
dc.title.journalZ Zagadnien Nauk Sądowych
dc.typeJournalArticle
dspace.entity.typeArticle