Hierarchical Multi Task Learning With CTC
share
In Automatic Speech Recognition, it is still challenging to learn useful intermediate representations when using of high-level (or abstract) target units such as words. Character or phoneme based systems tend to outperform word based systems as long as thousands of hours of training data are being used. In this paper, we first show how hierarchical multi-task training can encourage the formation of useful intermediate representations. We achieve this by performing Connectionist Temporal Classification at different levels of the network with targets of different granularity. Our model thus performs predictions in multiple scales of granularity for the same input. On the standard 300h Switchboard training setup, our hierarchical multi-task architecture exhibits improvements over single-task architectures with the same number of parameters. Our model obtains 14.0 Switchboard subset without any decoder or language model, outperforming the current state-of-the-art on acoustic-to-word models.
READ FULL TEXT