Autors: Ferreira, D.D., Leira, L., Mihaylova, P. T., Georgieva, P.
Title: Breaking Text-Based CAPTCHA with Sparse Convolutional Neural Networks
Keywords: Automated test, CAPTCHAs, Convolutional neural network, Inte

Abstract: CAPTCHA is an automated test designed to check if the user is human. Though other approaches are explored (such as object recognition), the text-based CAPTCHA is still the main test used by many web service providers, to separate human users from bots. In this paper, a sparse Convolutional Neural Network (CNN) to break text-based CAPTCHA is proposed. Unlike previous CNN solutions, which mainly use fine-tuning and transfer learning from pre-trained models, the proposed framework does not require a pre-trained model. The sparsity constraint deactivates connections between neurons in the CNN fully connected layers that leads to improved accuracy compared to the baseline approach. Visualization of the spatial distribution of neuron activity shed light on the internal learning and the effect of the sparsity constraint.

References

    Issue

    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 9th Iberian Conference on Pattern Recognition and Image Analysis, IbPRIA 2019, Madrid, Spain, vol. 11868 LNCS, pp. 404-415, 2019, Switzerland, Springer Nature, DOI 10.1007/978-3-030-31321-0_35

    Copyright Springer Nature Switzerland AG.

    Цитирания (Citation/s):
    1. Wang, Y., Wei, Y., Zhang, Y. et al. Few-shot learning in realistic settings for text CAPTCHA recognition. Neural Comput & Applic 35, 10751–10764 (2023). https://doi.org/10.1007/s00521-023-08262-0 - 2023 - в издания, индексирани в Scopus или Web of Science
    2. N. Zhang, M. Ebrahimi, W. Li and H. Chen, "A Generative Adversarial Learning Framework for Breaking Text-Based CAPTCHA in the Dark Web," 2020 IEEE International Conference on Intelligence and Security Informatics (ISI), Arlington, VA, USA, 2020, pp. 1-6, doi: 10.1109/ISI49825.2020.9280537. - 2020 - в издания, индексирани в Scopus или Web of Science

    Вид: пленарен доклад в международен форум, публикация в реферирано издание, индексирана в Scopus