A novel approach for image classification based on extreme learning machine

Image classification is an important task in content-based image retrieval, which can be regarded as an intermediate component to handle large-scale image datasets for improving the accuracy of image retrieval. Traditional image classification methods generally utilize Support Vector Machines (SVM) as image classifier. However, there are several drawbacks of using SVM, such as the high computational cost and large number of parameters to be optimized. In this paper we propose an Extreme Learning Machine (ELM) based Multi-modality Classifier Combination Framework (MCCF) to improve the accuracy of image classification. In this framework: (i) three ELM classifiers are trained by exploring three kinds of visual features respectively, (ii) a probability based fusion method is then proposed to combine the prediction results of each ELM classifier. Experiments on the widely used TRECVID datasets demonstrate that our approach can effectively improve the accuracy of image classification and achieve performance at extremely high speed.