Related Experiment Video
Updated: Oct 23, 2025
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
858
Towards open-set touchless palmprint recognition via weight-based meta metric learning
Huikai Shao1, Dexing Zhong1,2,3
1School of Automation Science and Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi 710049, China.
Pattern Recognition
|August 17, 2021
Summary
Touchless palmprint recognition is crucial for hygiene, especially during COVID-19. A new Weight-based Meta Metric Learning (W2ML) method enhances open-set recognition accuracy and efficiency for contactless biometrics.
Area of Science:
- Biometrics
- Computer Vision
- Machine Learning
Background:
- Touchless biometrics, particularly palmprint recognition, has gained importance due to hygiene concerns during the COVID-19 pandemic.
- Existing methods primarily focus on close-set scenarios, limiting their applicability in real-world, open-set conditions.
- The need for accurate and robust recognition systems that can handle unseen data is critical.
Purpose of the Study:
- To propose a novel Weight-based Meta Metric Learning (W2ML) method for accurate open-set touchless palmprint recognition.
- To enhance the generalization ability of feature extractors through meta-learning.
- To improve the efficiency and robustness of palmprint recognition in unconstrained environments.
Main Methods:
- Developed a Weight-based Meta Metric Learning (W2ML) approach for open-set palmprint recognition.
- Employed a deep metric learning-based feature extractor trained in a meta-learning framework.
- Utilized random sampling to create support and query sets, combined into meta sets for distance constraints.
- Incorporated hard sample mining and weighting to optimize meta set selection for improved efficiency.
Main Results:
- The W2ML method demonstrated superior robustness and efficiency in open-set touchless palmprint recognition compared to state-of-the-art techniques.
- Achieved accuracy improvements of up to 9.11% on benchmark datasets.
- Reduced the Equal Error Rate (EER) by up to 2.97%, indicating enhanced identification and verification performance.
- Validated on four palmprint benchmarks comprising fourteen constrained and unconstrained datasets.
Conclusions:
- The proposed W2ML method effectively addresses the challenges of open-set touchless palmprint recognition.
- Meta-learning and strategic sample selection significantly improve feature generalization and recognition performance.
- W2ML offers a promising solution for secure and hygienic biometric identification in diverse, real-world scenarios.
Related Concept Videos
00:57Weighted Mean
5.8K
While taking the arithmetic, geometric, or harmonic mean of a sample data set, equal importance is assigned to all the data points. However, all the values may not always be equally important in some data sets. An intrinsic bias might make it more important to give more weightage to specific values over others.
For example, consider the number of goals scored in the matches of a tournament. While computing the average number of goals scored in the tournament, it may be more important to...
For example, consider the number of goals scored in the matches of a tournament. While computing the average number of goals scored in the tournament, it may be more important to...
5.8K
01:03IR Frequency Region: Fingerprint Region
1.4K
IR spectra are divided into two main regions: the diagnostic region and the fingerprint region. The diagnostic region of the spectrum lies above 1500 cm−1. The absorptions resulting from single-bond vibrations of the N–H, C–H, and O–H stretch at higher wavenumbers and appear on the left side of the spectrum. The stretching absorptions of the C≡C and C≡N occur between 2100–2300 cm−1. In contrast, those arising from stretching absorptions of the...
1.4K