The CNNRF can be considered a cross-modal prediction model, which is a challenging research area that requires more attention because it is closely related to associative thinking and creativity.