Structured output layer with auxiliary targets for context-dependent acoustic modelling.

Pawel Swietojanski, Peter Bell, Steve Renals

Browse the full Interspeech paper archive.