Due to the uncertainty of clothing styles and the overwhelming amount of clothing product information, consumers often face difficulty in making choices when shopping online, and the results of their choices are often unsatisfactory. In response to this situation, this study proposes a women's shirt style recognition method based on the VGG16 convolutional neural network model. A sample library containing six categories of women's shirt style images was created. The VGG16 convolutional neural network model was established using transfer learning methods and underwent corresponding training. After constructing the appropriate dataset, this network model was used to classify and recognise the styles of women's shirt samples in the dataset. Finally, the calculated recognition classification accuracy of this model was found to be 0.86. The VGG16-CNN's ability to recognise women's shirt styles demonstrates good feasibility, providing data references for women's shirt style recognition.

women's shirt, style recognition, convolutional neural network, VGG16