Diffusion-based Generative Speech Source Separation

Scheibler, Robin; Ji, Youna; Chung, Soo-Whan; Choe, Soyeon; Choi, Minseok

doi:10.48550/arxiv.2210.17327

Cited by 1 publication

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Source separation. Source separation is a task that aims to recover speech of interest from a mixed signal [94].…”

Section: Miscallenous Audio Tasksmentioning

confidence: 99%

“…However, using STFT in audio separation is not sensitive to phase and needs high-frequency resolution, making it low accuracy of reconstruction and time-consuming. To overcome these problems, other time-domain approaches are proposed, including independent component analysis (ICA) [12], non-negative matrix factorization(NMF) [131].DiffSep [94] is the first diffusion-based method for single-channel speech separation.…”

Section: Miscallenous Audio Tasksmentioning

confidence: 99%

“…Specifically, DiffSep [94] trains a SDE model that converts the separated signal to the mixtures, thus the reverse process can separate individual source from the mixed signal. DiffSep [94] can also be applied for audio denoising since noise can be seen as a type of extra source.…”

Section: Miscallenous Audio Tasksmentioning

confidence: 99%

See 2 more Smart Citations

Audio Diffusion Model for Speech Synthesis: A Survey on Text To Speech and Speech Enhancement in Generative AI

Zhang¹,

Zhang²,

Zheng³

et al. 2023

Preprint

View full text Add to dashboard Cite

Generative AI has demonstrated impressive performance in various fields, among which speech synthesis is an interesting direction.With the diffusion model as the most popular generative model, numerous works have attempted two active tasks: text to speech and speech enhancement. This work conducts a survey on audio diffusion model, which is complementary to existing surveys that either lack the recent progress of diffusion-based speech synthesis or highlight an overall picture of applying diffusion model in multiple fields. Specifically, this work first briefly introduces the background of audio and diffusion model. As for the text-to-speech task, we divide the methods into three categories based on the stage where diffusion model is adopted: acoustic model, vocoder and end-to-end framework. Moreover, we categorize various speech enhancement tasks by either certain signals are removed or added into the input speech. Comparisons of experimental results and discussions are also covered in this survey.

show abstract

“…Source separation. Source separation is a task that aims to recover speech of interest from a mixed signal [94].…”

Section: Miscallenous Audio Tasksmentioning

confidence: 99%

Section: Miscallenous Audio Tasksmentioning

confidence: 99%