A fusion neural network classifier for image classification

摘要

Neural networks have been commonly used for image classification problems by fusing input features extracted from multiple MPEG-7 descriptors. It is because they can provide better performance than those extracted from single descriptor. However the input feature dimension can be various according to MPEG-7 descriptors. Usually input features with large dimension are dominant over those with small dimension for generating outputs of the neural networks, even though their contribution to output is almost same. In order to solve the problem, we propose a fusion neural network classifier which divides each descriptor by the number of its input features. And we consider the importance of the input features in each descriptor during training the classifier. In the experimental section, we showed the analysis of our method and compared the performance of sports image classification with conventional neural network classifier, using six classes of sports images collected on the Internet.