Facial tracking using artificial neural networks
K. Naga Sujatha, P. Raja Rajeswari, S. Purusothaman · 2005
A region-based method for facial tracking is proposed. The method fully utilizes the facial information of temporal motion and spatial luminance. The dominant motion of the tracked facial object is computed. Using this result, the object template is warped to generate a prediction template. A method is proposed to modify the prediction; it incorporates an artificial neural network (ANN) with a backpropagation algorithm (BPA). A decision approach, with a threshold, is used to detect if there is any change in the object in successive frames. The accuracy of the result depends on the number of nodes in the hidden layer and the learning factor. The number of nodes in the hidden layer is 10, and the learning factor is 1. The performance of the algorithm, in reconstructing the tracked object, is about 96.5%, in terms of reduced time and quality of reconstruction.