HMM-based DNN?

[Hanzhang X]

In Zen et al. (2013), is the model still HMM based? Is only the decision-tree of HMM replaced by a DNN, aiming at representing context better? But it seems to me that a DNN-based model is completely different from a HMM-based one.

[Simon]

Good – you have correctly understood that the DNN is not replacing the HMM, but rather it is replacing the regression tree that is used to cluster (tie) the HMM parameters. In current DNN systems, we still need a separate model of the sequence (and in particular, something that divides phones into sub-phonetic units and models the durations of those sub-phone units): that is why there is still an HMM involved, and the HMM state is the sub-phone unit.

In the next lecture I will spell out the relationship between the two approaches in more detail.

[Author]

Posts