Deepfake Speech Detection: A Spectrogram Analysis

Simple item page
dc.contributor.authorFirc, Antoncs
dc.contributor.authorMalinka, Kamilcs
dc.contributor.authorHanáček, Petrcs
dc.date.issued2024-04-08cs
dc.description.abstractThe current voice biometric systems have no natural mechanics to defend against deepfake spoofing attacks. Thus, supporting these systems with a deepfake detection solution is necessary. One of the latest approaches to deepfake speech detection is representing speech as a spectrogram and using it as an input for a deep neural network. This work thus analyzes the feasibility of different spectrograms for deepfake speech detection. We compare types of them regarding their performance, hardware requirements, and speed. We show the majority of the spectrograms are feasible for deepfake detection. However, there is no general, correct answer to selecting the best spectrogram. As we demonstrate, different spectrograms are suitable for different needs.en
dc.description.abstractThe current voice biometric systems have no natural mechanics to defend against deepfake spoofing attacks. Thus, supporting these systems with a deepfake detection solution is necessary. One of the latest approaches to deepfake speech detection is representing speech as a spectrogram and using it as an input for a deep neural network. This work thus analyzes the feasibility of different spectrograms for deepfake speech detection. We compare types of them regarding their performance, hardware requirements, and speed. We show the majority of the spectrograms are feasible for deepfake detection. However, there is no general, correct answer to selecting the best spectrogram. As we demonstrate, different spectrograms are suitable for different needs.en
dc.formattextcs
dc.format.extent1312-1320cs
dc.format.mimetypeapplication/pdfcs
dc.identifier.citationProceedings of the ACM Symposium on Applied Computing. 2024, p. 1312-1320.en
dc.identifier.doi10.1145/3605098.3635911cs
dc.identifier.isbn979-8-4007-0243-3cs
dc.identifier.orcid0000-0002-4717-1910cs
dc.identifier.orcid0000-0002-9009-2193cs
dc.identifier.orcid0000-0001-5507-0768cs
dc.identifier.other188028cs
dc.identifier.researcheridHJP-8074-2023cs
dc.identifier.researcheridAAB-5046-2022cs
dc.identifier.scopus57699371300cs
dc.identifier.scopus24824985000cs
dc.identifier.scopus6508388287cs
dc.identifier.urihttp://hdl.handle.net/11012/252867
dc.language.isoencs
dc.publisherAssociation for Computing Machinerycs
dc.relation.ispartofProceedings of the ACM Symposium on Applied Computingcs
dc.relation.urihttps://dl.acm.org/doi/10.1145/3605098.3635911cs
dc.rightsCreative Commons Attribution 4.0 Internationalcs
dc.rights.accessopenAccesscs
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/cs
dc.subjectDeepfakeen
dc.subjectSpeechen
dc.subjectImage-baseden
dc.subjectDeepfake Detectionen
dc.subjectSpectrogramen
dc.subjectDeepfake
dc.subjectSpeech
dc.subjectImage-based
dc.subjectDeepfake Detection
dc.subjectSpectrogram
dc.titleDeepfake Speech Detection: A Spectrogram Analysisen
dc.title.alternativeDeepfake Speech Detection: A Spectrogram Analysisen
dc.type.driverconferenceObjecten
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen
sync.item.dbidVAV-188028en
sync.item.dbtypeVAVen
sync.item.insts2025.10.14 14:13:18en
sync.item.modts2025.10.14 10:08:32en
thesis.grantorVysoké učení technické v Brně. Fakulta informačních technologií. Ústav inteligentních systémůcs

Files

Original bundle

Now showing 1 - 1 of 1
Thumbnail Image
Name:
3605098.3635911.pdf
Size:
2.81 MB
Format:
Adobe Portable Document Format
Description:
file 3605098.3635911.pdf
Download

Collections

Ústav inteligentních systémů