Using a Selective Ensemble Support Vector Machine to Fuse Multimodal Features for Human Action Recognition

Figure 3

DMM-LBP feature extraction algorithm flow.