HMM-based text-to-articulatory-movement prediction and analysis of critical articulators

In this paper we present a method to predict the movement of a speaker's mouth from text input using hidden Markov models (HMM). We have used a corpus of human articulatory movements, recorded by electromagnetic articulography (EMA), to train HMMs. To predict articulatory movements from text, a suitable model sequence is selected and the maximum-likelihood parameter generation (MLPG) algorithm is used to generate output articulatory trajectories. In our experiments, we find that fully context-dependent models outperform monophone and quinphone models, achieving an average

• Type: Journal
• Year: 2010
• Where: INTERSPEECH
• Authors: Zhen-Hua Ling, Korin Richmond, Junichi Yamagishi
• Keywords: Articulatory Movements, Human Articulatory Movements, INTERSPEECH 2010, Output Articulatory Trajectories, Signal Processing