Fig. 5

Position attention module. Position attention module effectively learns the dependency relationships between different positions, thereby improving the accuracy of feature representation. The Position Attention Module calculates the similarity between each position and other positions, and then weights them based on their similarity to aggregate information from different positions. Then, these weighted and aggregated feature information is sent to a fully connected layer for further reconstruction, which enhances the representation ability of features. Ultimately, the high-level features output by the Position Attention Module will contain more information about the relationship between different positions in the input features, which can help improve the model’s performance and generalization ability