Blurring Scene Recognition in Short Video

In order to solve the problems of blur and jitter in short video scenes, this paper proposes a multi-scale fusion network based on VGGNet. First, a fuzzy image data set is obtained by performing different blur processing on the clear image data set. Then, the fuzzy feature is extracted from the blurred image by using the fused depth network, and compared with the clear image extraction feature for back propagation training. Finally, the classification of fuzzy scene recognition is performed. The top3 of the Charades short video dataset has achieved 78.9% of the results, which proves that the proposed method has a good performance in blurring scene recognition.