An application of classifier combination methods in hand gesture recognition

Hand gesture recognition is a topic in artificial intelligence and computer vision with the goal to automatically interpret human hand gestures via some algorithms. Notice that it is a difficult classification task for which only one simple classifier cannot achieve satisfactory performance; several classifier combination techniques are employed in this paper to handle this specific problem. Based on some related data at hand, AdaBoost and rotation forest are seen to behave significantly better than all the other considered algorithms, especially a classification tree. By investigating the bias-variance decompositions of error for all the compared algorithms, the success of AdaBoost and rotation forest can be attributed to the fact that each of them simultaneously reduces the bias and variance terms of a SingleTree's error to a large extent. Meanwhile, kappa-error diagrams are utilized to study the diversity-accuracy patterns of the constructed ensemble classifiers in a visual manner.