Spatial Cross-Attention for Transformer-Based Image Captioning | IEEE Conference Publication | IEEE Xplore