Block diagram of the proposed SUPP Network. The network is composed of two UPPNet blocks and four prediction heads. The first UPPNet block forms coarse predictions, and high-resolution feature maps that are forwarded to the second block (dashed arrow). The coarse predictions in intermediate outputs Cont 1 and Seg 1 (first pseudo-continuum and segmentation outputs, respectively) are forwarded to the second block. The second block forms the final predictions at Cont 2 and Seg 2 outputs.