SpringerLink
Log in

Active Learning and Transfer Learning for Document Segmentation

  • Published:
Programming and Computer Software Aims and scope Submit manuscript

Abstract

In this paper, we investigate the effectiveness of classical approaches to active learning in the problem of document segmentation with the aim of reducing the size of the training sample. A modified approach to selection of document images for labeling and subsequent model training is presented. The results of active learning are compared to those of transfer learning on fully labeled data. The paper also investigates how the problem domain of a training set, on which a model is initialized for transfer learning, affects the subsequent uptraining of the model.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.
Fig. 5.

Similar content being viewed by others

Notes

  1. https://pytorch.org.

REFERENCES

  1. Settles, B., Active learning literature survey, Technical report no. 1648, University of Wisconsin-Madison, 2009.

  2. Scheffer, T., Decomain, C., and Wrobel, S., Active hidden Markov models for information extraction, Proc. Int. Symp. Intelligent Data Analysis, 2001, pp. 309–318.

  3. Dagan, I. and Engelson, S., Committee-based sampling for training probabilistic classifiers, Proc. 12th Int. Conf. Machine Learning, 1995, pp. 150–157.

  4. Culotta, A. and McCallum, A., Reducing labeling effort for structured prediction tasks, Proc. 20th Natl. Conf. Artificial Intelligence, 2005, pp. 746–751.

  5. Brust, C., Käding, C., and Denzler, J., Active learning for deep object detection, Proc. 14th Int. Jt. Conf. Computer Vision, Imaging and Computer Graphics Theory and Applications, 2019, pp. 181–190.

  6. Kao, C., Lee, T., et al., Localization-aware active learning for object detection, Proc. 14th Asian Conf. Computer Vision, 2018, pp. 506–522.

  7. Roy, S., Unmesh, A., and Namboodiri, V., Deep active learning for object detection, Proc. 29th British Machine Vision Conf., 2018.

  8. Aghdam, H., Gonzalez-Garcia, A., et al., Active learning for deep detection neural networks, Proc. 17th IEEE/CVF Int. Conf. Computer Vision, 2019, pp. 3671–3679.

  9. Lv, X., Duan, F., et al., Deep active learning for surface defect detection, Sensors, 2020, vol. 20, no. 6.

  10. Lin, T., Maire, M., et al., Microsoft COCO: Common objects in context, Lect. Notes Comput. Sci., 2014, vol. 8693, pp. 740–755.

    Article  Google Scholar 

  11. Zhong, X., Tang, J., and Yepes, A., PubLayNet: Largest dataset ever for document layout analysis, Proc. Int. Conf. Document Analysis and Recognition (ICDAR), 2019, pp. 1015–1022.

  12. Belyaeva, O.V., Perminov, A.I., and Kozlov, I.S., Synthetic data usage for document segmentation models fine-tuning, Tr. Inst. Sist. Program. Ross. Akad. Nauk, 2020, vol. 32, no. 4, pp. 189–202. https://doi.org/10.15514/ISPRAS-2020-32(4)-14

  13. Shen, Z., Zhao, J., et al., OLALA: Object-level active learning for efficient document layout annotation, 2021.

  14. Ren, S., He, K., et al., Faster R-CNN: Towards real-time object detection with region proposal networks, Proc. 28th Int. Conf. Neural Information Processing Systems, 2015, pp. 91–99.

  15. He, K., Gkioxari, G., et al., Mask R-CNN, Proc. IEEE Int. Conf. Computer Vision (ICCV), 2017, pp. 2980–2988.

Download references

Author information

Authors and Affiliations

  1. Ivannikov Institute for System Programming, Russian Academy of Sciences, ul. Solzhenitsyna 25, 109004, Moscow, Russia

    D. M. Kiranov, M. A. Ryndin & I. S. Kozlov

  2. Moscow Institute of Physics and Technology, Institutskii per. 9, 141700, Dolgoprudnyi, Moscow oblast, Russia

    D. M. Kiranov

Authors
  1. D. M. Kiranov
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. M. A. Ryndin
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. I. S. Kozlov
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding authors

Correspondence to D. M. Kiranov, M. A. Ryndin or I. S. Kozlov.

Ethics declarations

The authors declare that they have no conflicts of interest.

Additional information

Translated by Yu. Kornienko

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kiranov, D.M., Ryndin, M.A. & Kozlov, I.S. Active Learning and Transfer Learning for Document Segmentation. Program Comput Soft 49, 566–573 (2023). https://doi.org/10.1134/S0361768823070046

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1134/S0361768823070046

Keywords:

Search

Navigation