MCCA: Multimodal clothing compatibility learning based on channel attention

In order to slove the problems of incomplete feature extraction of clothing images and difficult modeling of clothing compatibility, a model named ECA-RMCN (a multimodal clothing compatibility model based on channel attention) is proposed. The ECA-Net(a high-efficiency channel attention module) is introduced on the residual module of the feature extraction network CNN to enhance important features such as low-level and high-level clothing, and suppress invalid features; the combined loss function is used to deal with the problem of unbalanced between positive samples and negative samples of clothing to achieve better matching effect. Comparative experiments are performed on the public Polyvore dataset to validate whether the effective of the model. The experiment data illustrate that the algorithm is excelled than else algorithm in the compatibility prediction and matching performance of clothing, and has good application value.