Diffuse or Confuse: A Diffusion Deepfake Speech Dataset
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866929674676862976 |
|---|---|
| author | Firc, Anton Malinka, Kamil Hanáček, Petr |
| author_facet | Firc, Anton Malinka, Kamil Hanáček, Petr |
| contents | Advancements in artificial intelligence and machine learning have significantly improved synthetic speech generation. This paper explores diffusion models, a novel method for creating realistic synthetic speech. We create a diffusion dataset using available tools and pretrained models. Additionally, this study assesses the quality of diffusion-generated deepfakes versus non-diffusion ones and their potential threat to current deepfake detection systems. Findings indicate that the detection of diffusion-based deepfakes is generally comparable to non-diffusion deepfakes, with some variability based on detector architecture. Re-vocoding with diffusion vocoders shows minimal impact, and the overall speech quality is comparable to non-diffusion methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_06796 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Diffuse or Confuse: A Diffusion Deepfake Speech Dataset Firc, Anton Malinka, Kamil Hanáček, Petr Cryptography and Security Artificial Intelligence Machine Learning Sound I.2.7 Advancements in artificial intelligence and machine learning have significantly improved synthetic speech generation. This paper explores diffusion models, a novel method for creating realistic synthetic speech. We create a diffusion dataset using available tools and pretrained models. Additionally, this study assesses the quality of diffusion-generated deepfakes versus non-diffusion ones and their potential threat to current deepfake detection systems. Findings indicate that the detection of diffusion-based deepfakes is generally comparable to non-diffusion deepfakes, with some variability based on detector architecture. Re-vocoding with diffusion vocoders shows minimal impact, and the overall speech quality is comparable to non-diffusion methods. |
| title | Diffuse or Confuse: A Diffusion Deepfake Speech Dataset |
| topic | Cryptography and Security Artificial Intelligence Machine Learning Sound I.2.7 |
| url | https://arxiv.org/abs/2410.06796 |