MMSR

Method

Combining Visual and Text Features", which is used to support the data statistics, main experimental results, ablation experimental results, and manual evaluation results of the generated interpretation quality in the paper. The paper experiment uses three publicly available multimodal metaphor datasets: Multi MET, MultiCMET, and MET Met. Due to the fact that the original image data is sourced from a third-party public dataset and involves third-party intellectual property rights, this data has not been repeatedly stored or publicly released as original image files.