Conference paper · 2021 · Open access

Improved Residual Networks for Image and Video Recognition

Ionuţ Cosmin Duţă, Li Liu, Fan Zhu, Ling Shao

Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition, 9415–9422 · Published 10 January 2021 · DOI 10.1109/icpr48806.2021.9412193

Open PDFView at the publisherCite this paper

In brief

“Improved Residual Networks for Image and Video Recognition” is a 2021 conference paper by Ionuţ Cosmin Duţă and 3 co-authors, published in Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition. OpenAlex counts 298 citations as of 18 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).

Citations
298
Most cited in
2024 (87)
Field-weighted impact
12.42
References
42
Authors
4

Abstract

Residual networks (ResNets) represent a powerful type of convolutional neural network (CNN) architecture, widely adopted and used in various tasks. In this work we propose an improved version of ResNets. Our proposed improvements address all three main components of a ResNet: the flow of information through the network layers, the residual building block, and the projection shortcut. We are able to show consistent improvements in accuracy and learning convergence over the baseline. For instance, on ImageNet dataset, using the ResNet with 50 layers, for top-1 accuracy we can report a 1.19% improvement over the baseline in one setting and around 2% boost in another. Importantly, these improvements are obtained without increasing the model complexity. Our proposed approach allows us to train extremely deep networks, while the baseline shows severe optimization issues. We report results on three tasks over six datasets: image classification (ImageNet, CIFAR-10 and CIFAR-100), object detection (COCO) and video action recognition (Kinetics-400 and Something-Something-v2). In the deep learning era, we establish a new milestone for the depth of a CNN. We successfully train a 404-layer deep CNN on the ImageNet dataset and a 3002-layer network on CIFAR-10 and CIFAR-100, while the baseline is not able to converge at such extreme depths. Code and models are publicly available at: https://github.com/iduta/iresnet.

Cite this paper

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

APA 7

Duţă, I. C., Liu, L., Zhu, F., & Shao, L. (2021). Improved Residual Networks for Image and Video Recognition. Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition, 9415–9422. https://doi.org/10.1109/icpr48806.2021.9412193

MLA 9

Duţă, Ionuţ Cosmin, et al. “Improved Residual Networks for Image and Video Recognition.” Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition, 2021, pp. 9415–9422. https://doi.org/10.1109/icpr48806.2021.9412193.

Chicago

Duţă, Ionuţ Cosmin, Li Liu, Fan Zhu, and Ling Shao. 2021. “Improved Residual Networks for Image and Video Recognition.” Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition: 9415–9422. https://doi.org/10.1109/icpr48806.2021.9412193.

Harvard

Duţă, I. C. et al. (2021) ‘Improved Residual Networks for Image and Video Recognition’, Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition, pp. 9415–9422. doi:10.1109/icpr48806.2021.9412193.

IEEE

I. C. Duţă, L. Liu, F. Zhu, and L. Shao, “Improved Residual Networks for Image and Video Recognition,” Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition, pp. 9415–9422, 2021, doi: 10.1109/icpr48806.2021.9412193.

BibTeX

@inproceedings{duta2021improved,
  title     = {{Improved Residual Networks for Image and Video Recognition}},
  author    = {Duţă, Ionuţ Cosmin and Liu, Li and Zhu, Fan and Shao, Ling},
  booktitle = {Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition},
  year      = {2021},
  pages     = {9415--9422},
  doi       = {10.1109/icpr48806.2021.9412193},
  url       = {https://doi.org/10.1109/icpr48806.2021.9412193}
}

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

Citations by year

YearCitationsShare of the best year
201911
202077
20211414
20223636
20235454
20248787
20256767
20263232

Its field-weighted citation impact is 12.42: it has been cited 12.42 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

  • ResNet
  • residual blocks
  • image classification
  • video action recognition
  • object detection
  • training convergence

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?

Duţă, I. C., Liu, L., Zhu, F., & Shao, L. (2021). Improved Residual Networks for Image and Video Recognition. Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition, 9415–9422. https://doi.org/10.1109/icpr48806.2021.9412193

How many times has this paper been cited?

298 times, according to OpenAlex on 18 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?

Ionuţ Cosmin Duţă, Li Liu, Fan Zhu, Ling Shao.

What is the DOI of this paper?

10.1109/icpr48806.2021.9412193, permanently at https://doi.org/10.1109/icpr48806.2021.9412193. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 18 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.