Full Contextual Attention for Multi-resolution Transformers in Semantic Segmentation | IEEE Conference Publication | IEEE Xplore