Transform and Tell: Entity-Aware News Image Captioning | IEEE Conference Publication | IEEE Xplore