ViTAR: Vision Transformer with Any Resolution

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Fan, Qihang, You, Quanzeng, Han, Xiaotian, Liu, Yongfei, Tao, Yunzhe, Huang, Huaibo, He, Ran, Yang, Hongxia
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917624608194560
author Fan, Qihang
You, Quanzeng
Han, Xiaotian
Liu, Yongfei
Tao, Yunzhe
Huang, Huaibo
He, Ran
Yang, Hongxia
author_facet Fan, Qihang
You, Quanzeng
Han, Xiaotian
Liu, Yongfei
Tao, Yunzhe
Huang, Huaibo
He, Ran
Yang, Hongxia
contents This paper tackles a significant challenge faced by Vision Transformers (ViTs): their constrained scalability across different image resolutions. Typically, ViTs experience a performance decline when processing resolutions different from those seen during training. Our work introduces two key innovations to address this issue. Firstly, we propose a novel module for dynamic resolution adjustment, designed with a single Transformer block, specifically to achieve highly efficient incremental token integration. Secondly, we introduce fuzzy positional encoding in the Vision Transformer to provide consistent positional awareness across multiple resolutions, thereby preventing overfitting to any single training resolution. Our resulting model, ViTAR (Vision Transformer with Any Resolution), demonstrates impressive adaptability, achieving 83.3\% top-1 accuracy at a 1120x1120 resolution and 80.4\% accuracy at a 4032x4032 resolution, all while reducing computational costs. ViTAR also shows strong performance in downstream tasks such as instance and semantic segmentation and can easily combined with self-supervised learning techniques like Masked AutoEncoder. Our work provides a cost-effective solution for enhancing the resolution scalability of ViTs, paving the way for more versatile and efficient high-resolution image processing.
format Preprint
id arxiv_https___arxiv_org_abs_2403_18361
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle ViTAR: Vision Transformer with Any Resolution
Fan, Qihang
You, Quanzeng
Han, Xiaotian
Liu, Yongfei
Tao, Yunzhe
Huang, Huaibo
He, Ran
Yang, Hongxia
Computer Vision and Pattern Recognition
This paper tackles a significant challenge faced by Vision Transformers (ViTs): their constrained scalability across different image resolutions. Typically, ViTs experience a performance decline when processing resolutions different from those seen during training. Our work introduces two key innovations to address this issue. Firstly, we propose a novel module for dynamic resolution adjustment, designed with a single Transformer block, specifically to achieve highly efficient incremental token integration. Secondly, we introduce fuzzy positional encoding in the Vision Transformer to provide consistent positional awareness across multiple resolutions, thereby preventing overfitting to any single training resolution. Our resulting model, ViTAR (Vision Transformer with Any Resolution), demonstrates impressive adaptability, achieving 83.3\% top-1 accuracy at a 1120x1120 resolution and 80.4\% accuracy at a 4032x4032 resolution, all while reducing computational costs. ViTAR also shows strong performance in downstream tasks such as instance and semantic segmentation and can easily combined with self-supervised learning techniques like Masked AutoEncoder. Our work provides a cost-effective solution for enhancing the resolution scalability of ViTs, paving the way for more versatile and efficient high-resolution image processing.
title ViTAR: Vision Transformer with Any Resolution
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.18361