Related Experiment Video
Updated: May 28, 2025
11:34
High-resolution, High-speed, Three-dimensional Video Imaging with Digital Fringe Projection Techniques
Published on: December 3, 2013
15.6K
Accurate Real-Time Live Face Detection Using Snapshot Spectral Imaging Method
Zhihai Wang1,2, Shuai Wang1, Weixing Yu1,2
1Key Laboratory of Spectral Imaging Technology, Xi'an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi'an 710119, China.
Sensors (Basel, Switzerland)
|February 13, 2025
Summary
This study introduces a novel live face detection (LFD) method using spectral imaging, achieving over 98% accuracy and under one-second detection times. This advanced LFD system offers a faster, more accurate alternative for identification verification.
Area of Science:
- Biometrics
- Computer Vision
- Spectroscopy
Background:
- Traditional facial recognition relies on 2D/3D images, limiting liveness detection.
- Existing methods face challenges in speed and accuracy for real-time identification verification.
Purpose of the Study:
- To develop a novel live face detection (LFD) method using snapshot spectral imaging.
- To enhance accuracy and speed in facial recognition systems for identification verification.
Main Methods:
- Utilized snapshot spectral imaging and computational spectral reconstruction (Tikhonov regularization).
- Employed Dlib face detection and Euclidean distance for face area extraction.
- Trained a convolutional neural network (CNN) on a spectral database for liveness prediction.
Main Results:
- Achieved spectral reconstruction fidelity over 99% for various samples.
- CNN model demonstrated over 98% accuracy in predicting liveness.
- System detection time consistently remained under one second, outperforming existing spectral imaging LFD methods.
Conclusions:
- The proposed spectral imaging LFD method offers a superior alternative for accurate, highly integrated video LFD applications.
- The system's speed and accuracy show significant potential for broader applications in recognition fields.
- Pixel-level liveness detection experiments validated the method's theoretical results and practical viability.
