Paper "Joint Inference of Objects and Scenes with Efficient Learning of Text-Object-Scene Relations" by Botao Wang et al. was accepted by IEEE Transactions on Multimedia (TMM).
2016-01-10 11