Continuous Speech Separation with Conformer

08/13/2020 ∙ by Sanyuan Chen, et al. ∙ 0 ∙ share

Continuous speech separation plays a vital role in complicated speech related tasks such as conversation transcription. The separation model extracts a single speaker signal from a mixed speech. In this paper, we use transformer and conformer in lieu of recurrent neural networks in the separation system, as we believe capturing global information with the self-attention based method is crucial for the speech separation. Evaluating on the LibriCSS dataset, the conformer separation model achieves state of the art results, with a relative 23.5 utterance-wise evaluation and a 15.4 evaluation.

READ FULL TEXT

Continuous speech separation: dataset and analysis

This paper describes a dataset and protocols for evaluating continuous s...
01/30/2020 ∙ by Zhuo Chen, et al. ∙ 0 ∙ share

read it
TransMask: A Compact and Fast Speech Separation Model Based on Transformer

Speech separation is an important problem in speech processing, which ta...
02/19/2021 ∙ by Zining Zhang, et al. ∙ 0 ∙ share

read it
Don't shoot butterfly with rifles: Multi-channel Continuous Speech Separation with Early Exit Transformer

With its strong modeling capacity that comes from a multi-head and multi...
10/23/2020 ∙ by Sanyuan Chen, et al. ∙ 0 ∙ share

read it
Multi-microphone Complex Spectral Mapping for Utterance-wise and Continuous Speaker Separation

We propose multi-microphone complex spectral mapping, a simple way of ap...
10/04/2020 ∙ by Zhong-Qiu Wang, et al. ∙ 0 ∙ share

read it
An End-to-end Architecture of Online Multi-channel Speech Separation

Multi-speaker speech recognition has been one of the keychallenges in co...
09/07/2020 ∙ by Jian Wu, et al. ∙ 0 ∙ share

read it
Dual-Path Transformer Network: Direct Context-Aware Modeling for End-to-End Monaural Speech Separation

The dominant speech separation models are based on complex recurrent or ...
07/28/2020 ∙ by Jingjing Chen, et al. ∙ 0 ∙ share

read it
SAGRNN: Self-Attentive Gated RNN for Binaural Speaker Separation with Interaural Cue Preservation

Most existing deep learning based binaural speaker separation systems fo...
09/02/2020 ∙ by Ke Tan, et al. ∙ 0 ∙ share

read it

Continuous Speech Separation with Conformer

Authors

Continuous speech separation: dataset and analysis

TransMask: A Compact and Fast Speech Separation Model Based on Transformer

Don't shoot butterfly with rifles: Multi-channel Continuous Speech Separation with Early Exit Transformer

Multi-microphone Complex Spectral Mapping for Utterance-wise and Continuous Speaker Separation

An End-to-end Architecture of Online Multi-channel Speech Separation

Dual-Path Transformer Network: Direct Context-Aware Modeling for End-to-End Monaural Speech Separation

SAGRNN: Self-Attentive Gated RNN for Binaural Speaker Separation with Interaural Cue Preservation

This week in AI