HIRI-ViT: Scaling Vision Transformer with High Resolution Inputs

arXiv:2403.11999v1 Announce Type: new
Abstract: The hybrid deep models of Vision Transformer (ViT) and Convolution Neural Network (CNN) have emerged as a powerful class of backbones for vision tasks. Scaling up the input resolution of such hybrid backbones naturally strengthes model capacity, but inevitably suffers from heavy computational cost that scales quadratically. Instead, we present a new hybrid backbone with HIgh-Resolution Inputs (namely HIRI-ViT), that upgrades prevalent four-stage ViT to five-stage ViT tailored for high-resolution inputs. HIRI-ViT is built upon …

abstract arxiv capacity class cnn computational convolution convolution neural network cost cs.cv cs.mm hybrid inputs network neural network scaling scaling up tasks transformer type vision vit