1. Home
2. Premium
3. IEEE ICASSP 2020 Virtual Conference May 2020
4. On End-To-End Multi-Channel Time Domain Speech Separation In Reverberant Environments

On End-To-End Multi-Channel Time Domain Speech Separation In Reverberant Environments

Collection:

IEEE ICASSP 2020 Virtual Conference May 2020

This paper introduces a new method for multi-channel time domain speech separation in reverberant environments. A fully-convolutional neural network structure has been used to directly separate speech from multiple microphone recordings, with no need of c

• IEEE MemberUS $11.00
• Society MemberUS $0.00
• IEEE Student MemberUS $11.00
• Non-IEEE MemberUS $15.00

Purchase

Videos in this product

On End-To-End Multi-Channel Time Domain Speech Separation In Reverberant Environments

00:13:24

0 views

This paper introduces a new method for multi-channel time domain speech separation in reverberant environments. A fully-convolutional neural network structure has been used to directly separate speech from multiple microphone recordings, with no need of c