Recognition and feedback of vowel utterance with a good mouth shape based on sensing Platysma muscle bulging

In public speaking, speakers are evaluated on verbal delivery and nonverbal delivery, and in particular, the mouth shape has an important role to support both of these. The mouth shape is mainly set during vowel utterance. We define the mouth shape, which can prompt the pronunciation of the speaker clearly and enrich the facial expression, as a good mouth shape in this research. The authors assume that a good mouth shape can be inferred from the bulging of the platysma muscle in the neck. We aim to support vowel utterances with a good mouth shape, and propose a system to recognize them. Specifically, we measure the uplift of the platysma muscle with photoreflectors and apply a machine learning method to implement a system to judge whether vowel utterances are being performed with a good shape. We conduct an accuracy measurement experiment of the proposed system and report the result. Finally, we describe the application that provides feedback of vowel utterances with a good mouth shape.