Abstract:
The new algorithm for clothes-based human recognition in video is presented. The algorithm is based on random patches classification with a random forest. The main advantage of the algorithm is the fact that it does not rely on human mask; therefore, it is able to work with video with arbitrary complex background. Experimental results are reported for a test set, which was acquired using authors' video surveillance system.
Keywords:video-based human recognition; machine learning; random forest; background subtraction.