Semantic Compositional Networks for Visual Captioning

CVPR 2017 Zhe GanChuang GanXiaodong HeYunchen PuKenneth TranJianfeng GaoLawrence CarinLi Deng

A Semantic Compositional Network (SCN) is developed for image captioning, in which semantic concepts (i.e., tags) are detected from the image, and the probability of each tag is used to compose the parameters in a long short-term memory (LSTM) network. The SCN extends each weight matrix of the LSTM to an ensemble of tag-dependent weight matrices... (read more)

PDF Abstract

Evaluation results from the paper


  Submit results from this paper to get state-of-the-art GitHub badges and help community compare results to other papers.