Human action recognition by bagging data dependent representation

Traditional methods based on bag-of-word representation are easily affected by noise, and they also cannot handle the problem when a test distribution differs from the training distribution. In this paper, we propose a novel method for human action recognition by bagging data dependent representation. Different with traditional methods, the proposed method represents each video by several histograms. These histograms are obtained by bagging according to an estimated prior several times in both training and testing. The data dependent property of our method depends on the prior which reflects the training distribution. There are two advantages of the proposed method. First, it alleviates the distribution difference between training set and test set. Second, the bagging operation reduces noise and improves the performance significantly. Experimental results show the effectiveness of the proposed method.