1 paper
Di Lu, Spencer Whitehead, Lifu Huang +2
Current image captioning approaches generate descriptions which lack specific information, such as named entities that are involved in the images. In this paper we propose a new ta…