Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 01023 | |
Number of page(s) | 6 | |
Section | Traffic Prediction and Analysis | |
DOI | https://doi.org/10.1051/itmconf/20257001023 | |
Published online | 23 January 2025 |
Federated Learning Applications in Fingerprint and Finger Vein Recognition
Computer Science, University of California, Davis, 95618 California, U.S.
Corresponding author: yowwang@ucdavis.edu
Fingerprints and finger veins are widely used in security identification in many fields due to their uniqueness and identifiability. However, their privacy issues are often criticized. This article summarizes several approaches that combine federated learning with fingerprint and finger vein recognition to solve privacy issues. One of the frameworks for fingerprint recognition, Federated Learning-Fingerprint Recognition, uses sparse representation techniques such as the Discrete Cosine Transform for data preprocessing. The framework also references the ResNet18 model and reservoir sampling so that each client can participate in training fairly. As for finger vein recognition, the Federated Learning-based Finger Vein authentication framework allows clients to share model weights to solve the data island problem and divide client data into shared and personalized parts to ensure privacy. This paper also points out its challenges, such as poor interpretability and applicability, and provides optimization solutions. For example, the interpretability issue can be solved by implementing an expert system. The expert system uses its robust knowledge base and inference engine to track model behavior and derive reasonable explanations. Transfer learning can also eliminate the applicability issue. It transfers the knowledge gained from training clients with concentrated data to clients with sparse data. In summary, this article comprehensively reviews the methods of federated learning in fingerprint and finger vein, respectively, and discusses the shortcomings and prospects.
© The Authors, published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.