Hidden Markov model based gesture recognition
Date of Issue2010
School of Mechanical and Aerospace Engineering
It is that extending more stable and more practicable algorithm in human robot interaction (HRI) for the increasing service robot market, however with less cost in consumer-able life is preferable. Hidden Markov Model performs well in conducting the recognition and in saving cost in device and program preparation. Hidden Markov Model has its special characteristics which are that the uncertain information in the modeled object is taken into consideration, which contributes to the high accuracy and low error rate during the recognition. Moreover, uncertain human neural information takes part in and affects the human gesture and intention meaning.A model which can involve this information can capture more information to the final recognition.
Nanyang Technological University