Mar 20, 2023 · This paper presents Cocktail HuBERT, a self-supervised learning framework that generalizes to mixture speech using a masked pseudo source separation objective.

Cocktail HuBERT outper- forms state-of-the-art results with 69% lower WER on multi-speaker. ASR, 31% lower DER on diarization, and is competitive on single- and ...

This paper presents Cocktail HuBERT, a self-supervised learning framework that generalizes to mixture speech using a masked pseudo source separation objective.

This paper presents Cocktail HuBERT, a self-supervised learning framework that generalizes to mixture speech using a masked pseudo source ...

Jul 16, 2024 · Cocktail Hubert: Generalized Self-Supervised Pre-Training for Mixture and Single-Source Speech. Maryam Fazel-Zarandi, Wei-Ning Hsu. 2023, IEEE ...

This paper presents Cocktail HuBERT, a self-supervised learning framework that generalizes to mixture speech using a masked pseudo source separation objective.

Mar 21, 2023 · Cocktail HuBERT: Generalized Self-Supervised Pre-training for Mixture and Single-Source Speech. arxiv.org. 77 · Like Comment. Share. Copy

Jul 3, 2024 · This work presents SA-WavLM, a novel pre-trained model for mixture speech. Specifically, SA-WavLM follows an “extract-merge-predict” pipeline.

May 5, 2024 · Self-supervised pre-trained speech models were shown effective for various downstream speech processing tasks.

This paper presents Cocktail HuBERT, a self-supervised learning framework that generalizes to mixture speech using a masked pseudo source separation objective.