Residual-block neural network with auxiliary speaker embeddings from clean recordings achieves 4.79 dB SDR, 8.44 dB SAR and 7.11 dB SIR on VoxCeleb for single-channel two-speaker separation.
A basic residual block contains two convolutional lay- ers, where batch normalisation [24] followed by a rectified lin- ear unit (ReLU) are applied between the convolutional layers
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Single-Channel Speech Separation with Auxiliary Speaker Embeddings
Residual-block neural network with auxiliary speaker embeddings from clean recordings achieves 4.79 dB SDR, 8.44 dB SAR and 7.11 dB SIR on VoxCeleb for single-channel two-speaker separation.