Video genre categorization using Support Vector Machines

In this paper, classifying and indexing video genres using Support Vector Machines (SVMs) are based on only audio features. In fact, those segmentation parameters are extracted at block levels, which have a major benefit by capturing local temporal information. The main contribution of our study is to present a powerful combination between the two employed audio descriptors Mel Frequency Cepstral Coefficients (MFCC) and signal energy in order to classify three common video genres: several sports analysis and matches, both studio and fields news scenes over and above various multi-speaker and multiinstruments music clips. Validation of this approach was carried out on over 6 hours of video span token from YouTube and yielding a classification accuracy of 99.83%. Finally we discuss SVM kernels performance on our proposed dataset.