Gesture recognition from image motion based on subspace method and HMM
説明
This paper proposes a method to recognize human gestures from an image sequence based on Hidden Markov Model and Karhunen-Lo`eve Transform. As our method uses the motion vector field of the scene for recognition, it is robust for variety in the background of the scene and it doesn't require the users to wear a sensor or a marker. The motion vector field of the scene is projected to an eigen-subspace for data compression and is used as the input symbols for the HMM. View-based methods generally fail when the user translates. Our method is robust for recognition to the user's translation because the recognition window automatically fits the user by tracking the user's face.