Conference paper · 2023 · Open access

Token-Label Alignment for Vision Transformers

Xiao Han, Wenzhao Zheng, Zheng Dong Zhu, Jie Zhou, Jiwen Lu

IEEE/CVF International Conference on Computer Vision (ICCV), 5472–5481 · Published 1 October 2023 · DOI 10.1109/iccv51070.2023.00506

Open PDFView at the publisherCite this paper

In brief

“Token-Label Alignment for Vision Transformers” is a 2023 conference paper by Xiao Han and 4 co-authors, published in IEEE/CVF International Conference on Computer Vision (ICCV). OpenAlex counts 9 citations as of 22 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).

Citations
9
Most cited in
2024 (6)
Field-weighted impact
0.52
References
57
Authors
5

Abstract

Data mixing strategies (e.g., CutMix) have shown the ability to greatly improve the performance of convolutional neural networks (CNNs). They mix two images as inputs for training and assign them with a mixed label with the same ratio. While they are shown effective for vision transformers (ViTs), we identify a token fluctuation phenomenon that has suppressed the potential of data mixing strategies. We empirically observe that the contributions of input tokens fluctuate as forward propagating, which might induce a different mixing ratio in the output tokens. The training target computed by the original data mixing strategy can thus be inaccurate, resulting in less effective training. To address this, we propose a token-label alignment (TL-Align) method to trace the correspondence between transformed tokens and the original tokens to maintain a label for each to-ken. We reuse the computed attention at each layer for efficient token-label alignment, introducing only negligible additional training costs. Extensive experiments demonstrate that our method improves the performance of ViTs on image classification, semantic segmentation, objective detection, and transfer learning tasks. Code is available at: https://github.com/Euphoria16/TL-Align.

Cite this paper

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

APA 7

Han, X., Zheng, W., Zhu, Z. D., Zhou, J., & Lu, J. (2023). Token-Label Alignment for Vision Transformers. IEEE/CVF International Conference on Computer Vision (ICCV), 5472–5481. https://doi.org/10.1109/iccv51070.2023.00506

MLA 9

Han, Xiao, et al. “Token-Label Alignment for Vision Transformers.” IEEE/CVF International Conference on Computer Vision (ICCV), 2023, pp. 5472–5481. https://doi.org/10.1109/iccv51070.2023.00506.

Chicago

Han, Xiao, Wenzhao Zheng, Zheng Dong Zhu, Jie Zhou, and Jiwen Lu. 2023. “Token-Label Alignment for Vision Transformers.” IEEE/CVF International Conference on Computer Vision (ICCV): 5472–5481. https://doi.org/10.1109/iccv51070.2023.00506.

Harvard

Han, X. et al. (2023) ‘Token-Label Alignment for Vision Transformers’, IEEE/CVF International Conference on Computer Vision (ICCV), pp. 5472–5481. doi:10.1109/iccv51070.2023.00506.

IEEE

X. Han, W. Zheng, Z. D. Zhu, J. Zhou, and J. Lu, “Token-Label Alignment for Vision Transformers,” IEEE/CVF International Conference on Computer Vision (ICCV), pp. 5472–5481, 2023, doi: 10.1109/iccv51070.2023.00506.

BibTeX

@inproceedings{han2023token,
  title     = {{Token-Label Alignment for Vision Transformers}},
  author    = {Han, Xiao and Zheng, Wenzhao and Zhu, Zheng Dong and Zhou, Jie and Lu, Jiwen},
  booktitle = {IEEE/CVF International Conference on Computer Vision (ICCV)},
  year      = {2023},
  pages     = {5472--5481},
  doi       = {10.1109/iccv51070.2023.00506},
  url       = {https://doi.org/10.1109/iccv51070.2023.00506}
}

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

Citations by year

YearCitationsShare of the best year
202466
202511
202622

Its field-weighted citation impact is 0.52: it has been cited 0.52 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
  • data mixing
  • CutMix
  • label assignment
  • image classification
  • semantic segmentation
  • object detection
  • transfer learning

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?

Han, X., Zheng, W., Zhu, Z. D., Zhou, J., & Lu, J. (2023). Token-Label Alignment for Vision Transformers. IEEE/CVF International Conference on Computer Vision (ICCV), 5472–5481. https://doi.org/10.1109/iccv51070.2023.00506

How many times has this paper been cited?

9 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?

Xiao Han, Wenzhao Zheng, Zheng Dong Zhu, Jie Zhou, Jiwen Lu.

What is the DOI of this paper?

10.1109/iccv51070.2023.00506, permanently at https://doi.org/10.1109/iccv51070.2023.00506. 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.