Abstract. We combine a neural image captioner with a Rational Speech Acts (RSA) model to make a system that is pragmatically informative: its objective is ...
確定！ 回上一頁
