Edinburgh Research Archive

Performance of deep neural networks employing predicted durations and stacked bottleneck features for speech synthesis

Item Status

RESTRICTED ACCESS

Embargo End Date

Date