Can I read ViT-FOD: A Vision Transformer based Fine-grained Object Discriminator on EtoBox?

ViT-FOD: A Vision Transformer based Fine-grained Object Discriminator by Zhang, Zi-Chao; Chen, Zhen-Duo; Wang, Yongxin; Luo, Xin; Xu, Xin-Shun is a scholarly article available to read on EtoBox.

What is ViT-FOD: A Vision Transformer based Fine-grained Object Discriminator about?

Recently, several Vision Transformer (ViT) based methods have been proposed for Fine-Grained Visual Classification (FGVC).These methods significantly surpass existing CNN-based ones, demonstrating the effectiveness of ViT in FGVC tasks.However, there are some limitations when applying ViT directly to FGVC.First, ViT needs to split images into patches and calculate the attention of every pair, which may result in heavy redundant calculation and unsatisfying performance when handling fine-grained images with complex background and small objects.Second, a standard ViT only utilizes the class token in the final layer for classification, which is not enough to extract comprehensive fine-grained information. To address these issues, we propose a novel ViT based fine-grained object discriminator for FGVC tasks, ViT-FOD for short. Specifically, besides a ViT backbone, it further introduces three novel components, i.e, Attention Patch Combination (APC), Critical Regions Filter (CRF), and Complementary Tokens Integration (CTI). Thereinto, APC pieces informative patches from two images to generate a new image so that the redundant calculation can be reduced. CRF emphasizes tokens correspondin

Author

Zhang, Zi-Chao; Chen, Zhen-Duo; Wang, Yongxin; Luo, Xin; Xu, Xin-Shun

Published

2022

Language

EN