Conference paper · 2021 · Open access

Incorporating Convolution Designs into Visual Transformers

Kun Yuan, Shaopeng Guo, Ziwei Liu, Aojun Zhou, Fengwei Yu, Wei Wu

IEEE/CVF International Conference on Computer Vision (ICCV), 559–568 · Published 1 October 2021 · DOI 10.1109/iccv48922.2021.00062

Open PDFView at the publisherCite this paper

In brief

“Incorporating Convolution Designs into Visual Transformers” is a 2021 conference paper by Kun Yuan and 5 co-authors, published in IEEE/CVF International Conference on Computer Vision (ICCV). OpenAlex counts 771 citations as of 22 November 2026, which puts it in the top 1% of works published in 2021 by citations. It is open access: the full text is free to read at arXiv (Cornell University).

Citations
771
Most cited in
2023 (205)
Field-weighted impact
35.46
References
57
Authors
6

Abstract

Motivated by the success of Transformers in natural language processing (NLP) tasks, there emerge some attempts (e.g., ViT and DeiT) to apply Transformers to the vision domain. However, pure Transformer architectures often require a large amount of training data or extra supervision to obtain comparable performance with convolutional neural networks (CNNs). To overcome these limitations, we analyze the potential drawbacks when directly borrowing Transformer architectures from NLP. Then we propose a new Convolution-enhanced image Transformer (CeiT) which combines the advantages of CNNs in extracting low-level features, strengthening locality, and the advantages of Transformers in establishing long-range dependencies. Three modifications are made to the original Transformer: 1) instead of the straightforward tokenization from raw input images, we design an Image-to-Tokens (I2T) module that extracts patches from generated low-level features; 2) the feed-froward network in each encoder block is replaced with a Locally-enhanced Feed-Forward (LeFF) layer that promotes the correlation among neighboring tokens in the spatial dimension; 3) a Layer-wise Class token Attention (LCA) is attached at the top of the Transformer that utilizes the multi-level representations.Experimental results on ImageNet and seven down-stream tasks show the effectiveness and generalization ability of CeiT compared with previous Transformers and state- of-the-art CNNs, without requiring a large amount of training data and extra CNN teachers. Besides, CeiT models demonstrate better convergence with 3× fewer training iterations, which can reduce the training cost significantly1.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Yuan, K., Guo, S., Liu, Z., Zhou, A., Yu, F., & Wu, W. (2021). Incorporating Convolution Designs into Visual Transformers. IEEE/CVF International Conference on Computer Vision (ICCV), 559–568. https://doi.org/10.1109/iccv48922.2021.00062

MLA 9

Yuan, Kun, et al. “Incorporating Convolution Designs into Visual Transformers.” IEEE/CVF International Conference on Computer Vision (ICCV), 2021, pp. 559–568. https://doi.org/10.1109/iccv48922.2021.00062.

Chicago

Yuan, Kun, Shaopeng Guo, Ziwei Liu, Aojun Zhou, Fengwei Yu, and Wei Wu. 2021. “Incorporating Convolution Designs into Visual Transformers.” IEEE/CVF International Conference on Computer Vision (ICCV): 559–568. https://doi.org/10.1109/iccv48922.2021.00062.

Harvard

Yuan, K. et al. (2021) ‘Incorporating Convolution Designs into Visual Transformers’, IEEE/CVF International Conference on Computer Vision (ICCV), pp. 559–568. doi:10.1109/iccv48922.2021.00062.

IEEE

K. Yuan, S. Guo, Z. Liu, A. Zhou, F. Yu, and W. Wu, “Incorporating Convolution Designs into Visual Transformers,” IEEE/CVF International Conference on Computer Vision (ICCV), pp. 559–568, 2021, doi: 10.1109/iccv48922.2021.00062.

BibTeX

@inproceedings{yuan2021incorporating,
  title     = {{Incorporating Convolution Designs into Visual Transformers}},
  author    = {Yuan, Kun and Guo, Shaopeng and Liu, Ziwei and Zhou, Aojun and Yu, Fengwei and Wu, Wei},
  booktitle = {IEEE/CVF International Conference on Computer Vision (ICCV)},
  year      = {2021},
  pages     = {559--568},
  doi       = {10.1109/iccv48922.2021.00062},
  url       = {https://doi.org/10.1109/iccv48922.2021.00062}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Citations by year

YearCitationsShare of the best year
20213131
2022134134
2023205205
2024173173
2025151151
20267777

Its field-weighted citation impact is 35.46: it has been cited 35.46 times as often as the average work of the same field, type and year (1.00 is the world average).

Topic and keywords

Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications

  • vision transformer
  • image tokenization
  • low-level feature extraction
  • locality
  • training data efficiency
  • training cost reduction
  • convergence rate

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Yuan, K., Guo, S., Liu, Z., Zhou, A., Yu, F., & Wu, W. (2021). Incorporating Convolution Designs into Visual Transformers. IEEE/CVF International Conference on Computer Vision (ICCV), 559–568. https://doi.org/10.1109/iccv48922.2021.00062

How many times has this paper been cited?

771 times, according to OpenAlex on 22 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.

Who wrote this paper?

Kun Yuan, Shaopeng Guo, Ziwei Liu, Aojun Zhou, Fengwei Yu, Wei Wu.

What is the DOI of this paper?

10.1109/iccv48922.2021.00062, permanently at https://doi.org/10.1109/iccv48922.2021.00062. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 22 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.