A Terrorist Scene Recognition Method Based on Pseudobrain Network Model

Abstract

The invention discloses a terrorism-related scene recognition method based on a pseudo-brain network model. The present invention comprises the following steps: Step 1, obtain the probability value of the scene picture involving terrorism through the CNN image classifier; Step 2, obtain the probability value of the scene picture involving terrorism through the pseudo-brain network image classifier; Step 3, obtain the probability value of the scene picture involving terrorism by the CNN image classifier The classification results and the classification results of the pseudo-brain network image classifier are input to the neural network fusion model, and the scene event classification results are output, where terrorism-related scenes=1 and non-terrorism-related scenes=0. The invention combines the deep learning algorithm with the equivalent pseudo-brain network model. The deep learning network is used for machine vision image classification, and the pseudo-brain network model replaces the human brain to fuse advanced features of audiovisual information to realize environmental object perception and environmental cognition.

Description

technical field [0001] The invention relates to the field of image recognition and classification, in particular to a method for recognizing terrorism-related scenes based on a pseudo-brain network model. This method can be applied to identify terrorism-related and violence-related scenes and other fields. Background technique [0002] With the development of online video, more and more videos related to terrorism and violence are widely disseminated on the Internet. These harmful videos can do a lot of harm because videos are more inflammatory and confusing than other media. Therefore, strengthening the supervision of violent and terrorist online video content has become an urgent need, which is of great significance for ensuring the security of online video content. Moreover, the current research results on related aspects of this field are very scarce, and the content of the present invention can effectively expand the application of this aspect. [0003] Early classic...

AI Technical Summary

Problems solved by technology

However, sound data is not intuitive information, so a complex calculation model is required to imitate the perception characteristics of the human ear for processing, especially the presence of noise will lead to large limitations such as distortion

At present, due to the lack of advanced computing models and methods that are more consistent with human perception characteristics, the bionic fusion of two kinds of heterogeneous information, sound and image, has not been studied in depth.

Images