Pertti Palo A Review of Articulatory Speech Synthesis

Articulatory speech synthesis models the natural speech production process. As a speech synthesis method it is not among the best, when the quality of produced speech sounds is the main criterion. However, for studying speech production it is the most suitable method. This thesis reviews the literature on articulatory speech synthesis. The objective is to look at articulatory speech synthesis form the viewpoint of basic speech research. From this viewpoint, articulatory speech synthesis is at its most interesting, when fidelity to natural speech production process is considered a model's most important aspect. The thesis reviews methods for articulatory measurements, geometrical synthesis of articulation and synthesis of speech sounds. The work divides the measurement methods to static and dynamic methods according to whether they are capable of capturing movement or not. In surveying geometrical synthesis the text concentrates especially on movement parametrisation and trajectory generation. Central topics in speech sound synthesis are the transmission line method and a glimpse at methods for non-linear sound generation.