A method for identifying clothing style outline features based on attention convolutional neural network
A clothing style contour feature recognition method based on attention convolutional neural network was proposed to address the problems of complex feature recognition techniques and low accuracy in existing clothing style contour feature recognition methods.Firstly,data augmentation methods was used to classify clothing style silhouette type labels;Then,the contour feature distribution gradient was calculated and determined through the loss function;After that,a feature recognition model was constructed using convolutional neural networks;Finally,the attention mechanism module was introduced to recognize the silhouette features of clothing styles.The verification results show that compared with the clothing style pattern recognition method based on improved Resnet34 and the clothing style recognition method based on improved edge detection algorithm,the proposed method always has a higher accuracy in identifying complex patterns.The accuracy of pattern recognition for dress style samples can reach 99.1%,and the styles of jackets,pants,and short sleeves can also reach over 90%.The recognition effect of the method in this article is accurate and effective,and can be extended to the recognition of clothing style contour features in reality.
NETL
NSTL
万方数据
