A vowel and semivowel synthesizer for prompt and expressive communication

In face-to-face communication, we mainly use the sound of our voice. Sound has many roles. For example, it transmit information as language, the emotion, the personality of the vocalizers, and so on. Therefor voice disorders cause much trouble in communicating with others. The people suffering from voice disorders use voice synthesizers, but they are not able to transmit all of the sound information. Even if they can clearly produce sounds of each word, they cannot produce them promptly and emotionally. We are trying to develop voice synthesizers that support prompt and expressive communication. Engineering seeks to keep existing functions and to improve devices to cover such disadvantages. On the other hand, this paper focuses on establishing human communication that eliminates existing functions and improves devices in another direction. Even though our proposed voice synthesizer can only produce the sounds of vowels and semivowels, it represents the stress and accent of sounds with rich inflection by synchronization with hand action. This paper also reports some experiments on its ability. The experimental results prove that vowel identification exceeds 80% and that a continuous shift of the sound is better than a discrete change for human perception.