Overview of our model: A first part computes image embeddings, a second one computes sentence embeddings through a Recurrent CapsNet. The two parts are ...
確定! 回上一頁