# ResFormer: Scaling ViTs with Multi-Resolution Training

Source: [ResFormer: Scaling ViTs with Multi-Resolution Training](<https://www.sota2.com/research/paper/68fa6ef21e209131cea4313c>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-27T06:22:26.632Z

Authors: Rui Tian, Zuxuan Wu, Qi Dai, Han Hu, Yu Qiao, Yu-Gang Jiang

Published: 2022-12-01T18:57:20.000Z

## Abstract

Vision Transformers (ViTs) have achieved overwhelming success, yet they suffer from vulnerable resolution scalability, i.e., the performance drops drastically when presented with input resolutions that are unseen during training. We introduce, ResFormer, a framework that is built upon the seminal idea of multi-resolution training for improved performance on a wide spectrum of, mostly unseen, testing resolutions. In particular, ResFormer operates on replicated images of different resolutions and enforces a scale consistency loss to engage interactive information across different scales. More importantly, to alternate among varying resolutions effectively, especially novel ones in testing, we propose a global-local positional embedding strategy that changes smoothly conditioned on input sizes. We conduct extensive experiments for image classification on ImageNet. The results provide strong quantitative evidence that ResFormer has promising scaling abilities towards a wide range of resolutions. For instance, ResFormer-B-MR achieves a Top-1 accuracy of 75.86% and 81.72% when evaluated on relatively low and high resolutions respectively (i.e., 96 and 640), which are 48% and 7.49% better than DeiT-B. We also demonstrate, moreover, ResFormer is flexible and can be easily extended to semantic segmentation, object detection and video action recognition. Code is available at https://github.com/ruitian12/resformer.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Image Classification on ImageNet-1K (Top-1 Accuracy)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-top-1-accuracy.md>): Image Classification; ImageNet-1K; Top-1 Acc: 83.47.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2212.00776>)

[PDF](<https://arxiv.org/pdf/2212.00776.pdf>)
