Talking vs Non-Talking: A Vision Based Approach to Detect Human Speaking Mode

Human talking mode detection is an important issue in human-computer interaction. In this work, we propose a method for detecting human talking and non talking mode detection based on supervised machine learning approach. Visual lip information of human is considered as an important clue. Our goal is to develop a method for human talking and non talking mode detection in real time using supervised classification algorithm. We tested our experiment with a single speaker task and compared the results with the previous method. The results show that our approach can obtain a 98.00% accuracy and a fast executed time.