F0 Modeling In Hmm-Based Speech Synthesis System Using Deep Belief Network
This work addresses F0 modeling in speech synthesis for Bengali, representing an incremental improvement by applying existing DBN methods to a specific domain.
The paper tackled the problem of modeling F0 contours in HMM-based speech synthesis for Bengali language by employing Deep Belief Networks (DBNs), resulting in improved F0 contours as shown in objective and subjective tests compared to clustering tree techniques.
In recent years multilayer perceptrons (MLPs) with many hid- den layers Deep Neural Network (DNN) has performed sur- prisingly well in many speech tasks, i.e. speech recognition, speaker verification, speech synthesis etc. Although in the context of F0 modeling these techniques has not been ex- ploited properly. In this paper, Deep Belief Network (DBN), a class of DNN family has been employed and applied to model the F0 contour of synthesized speech which was generated by HMM-based speech synthesis system. The experiment was done on Bengali language. Several DBN-DNN architectures ranging from four to seven hidden layers and up to 200 hid- den units per hidden layer was presented and evaluated. The results were compared against clustering tree techniques pop- ularly found in statistical parametric speech synthesis. We show that from textual inputs DBN-DNN learns a high level structure which in turn improves F0 contour in terms of ob- jective and subjective tests.