Policy Learning-Based Image Captioning With Vision Transformer | IEEE Conference Publication | IEEE Xplore