会議名: 2007年映像情報メディア学会年次大会
開催地: 工学院大学(新宿キャンパス)
開催日: 2007/08/29 - 2007/08/31
It is well known that the contents of utterance can be recognized from lip motions. This paper proposes a method for noncontact command input interface by using three kinds of lip motion features. The experimental results with 10 commands show that these lip motions are useful for command recognition.