Multi-modal feature fusion Android malicious software detection method based on attention mechanism

The invention relates to the technical field of malicious software detection, and discloses an attention mechanism-based multi-modal feature fusion Android malicious software detection method, which comprises the following steps of: extracting target contents in a dex file of target software, writing the target contents into a new file, converting the new file into an RGB (Red, Green and Blue) image, and inputting the RGB image into a deep convolutional neural network to obtain a picture feature vector; extracting permission information in a list file of the target software, and processing the permission information by utilizing a natural language processing model to obtain a text feature vector; and performing feature fusion on the picture feature vector and the text feature vector by adopting a multi-head attention mechanism to obtain a feature fusion vector, and inputting the feature fusion vector into a full connection layer to classify target software. According to the method, information from the visual