Dynamic texture video classification using extreme learning machine

Recognition of complex dynamic texture is a difficult task and captures the attention of the computer vision community for several decades. Essentially the dynamic texture recognition is a multi-class classification problem that has become a real challenge for computer vision and machine learning techniques. Due to the reason that the dynamic textures lie in non-Euclidean manifold, existing classifier such as extreme learning machine cannot effectively deal with this problem. In this paper, we propose a new approach to tackle the dynamic texture recognition problem. First, we utilize the affinity propagation clustering technology to design a codebook, and then construct a soft coding feature to represent the whole dynamic texture sequence. This new coding strategy preserves spatial and temporal characteristics of dynamic texture. Finally, by evaluating the proposed approach on the DynTex dataset, we show the effectiveness of the proposed strategy.