- Biometric Identification and Security
- Optical Coherence Tomography Applications
- Digital Media Forensic Detection
- Photoacoustic and Ultrasonic Imaging
- EEG and Brain-Computer Interfaces
- Spectroscopy Techniques in Biomedical and Chemical Research
- AI in cancer detection
- Natural Language Processing Techniques
- Cell Image Analysis Techniques
- Generative Adversarial Networks and Image Synthesis
- Thermography and Photoacoustic Techniques
- Emotion and Mood Recognition
- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Tactile and Sensory Interactions
- Forensic Fingerprint Detection Methods
- Functional Brain Connectivity Studies
- Advanced Image Fusion Techniques
- Ultrasound Imaging and Elastography
- Advancements in Transdermal Drug Delivery
- Cutaneous Melanoma Detection and Management
- Retinal Imaging and Analysis
- Machine Learning and Algorithms
- Advanced Steganography and Watermarking Techniques
- Algorithms and Data Compression
Zhejiang University of Technology
2020-2024
Nanjing University of Aeronautics and Astronautics
2017
Shandong University of Science and Technology
2017
The research of external fingerprint collected by total internal reflection (TIR) has been carried out for decades and the optical coherence tomography (OCT) just begun. can be hardly affected finger surface status, due to its strong antiinterference antispoofing ability, which serve as a powerful supplement fingerprint. However, matching fingerprints acquired in different ways lead drop recognition accuracy differences quality, distortions, detection areas. Whether used replace direct...
Presentation attack detection (PAD) is a critical component of automated fingerprint recognition systems (AFRSs). However, existing PAD technologies based on optical coherence tomography (OCT) mainly rely local information, ignoring the global continuity and correlation physiological structures. Furthermore, lack appropriate presentation instruments (PAIs) that cater to unique OCT characteristics leads insufficient evaluation PAD. The identification features, including external (EF),...
Anti-spoofing ability is vital for fingerprint identification systems. Conventional scanning devices can only obtain information from the fingertip surfaces, and their performance susceptible to skin conditions presentation attacks (PAs). However, optical coherence tomography (OCT) scan subcutaneous tissue 3D structures, naturally enhancing its PA detection (PAD) perspective of hardware. Existing unsupervised PAD methods are based on image reconstruction. reconstruction error easily affected...
Optical coherence tomography (OCT) is a noninvasive high-resolution imaging technology that can accurately acquire the internal characteristics of tissues within few millimeters. Using OCT technology, fingerprint structure, which consistent with external fingerprints and sweat glands, be collected, leading to high anti-spoofing capabilities. In this paper, an method based on 3D convolutional neural network (CNN) proposed, considering spatial continuity biometrics in fingertips. Experiments...
External fingerprints (EFs) based only on epidermal information are vulnerable to spoofing attacks and non-ideal skin conditions. To solve such shortcomings, internal (IFs) collected using optical coherence tomography (OCT) have been proposed widely researched. However, the development of IF is limited by lack in-depth researches EF-IF interoperability, which partially caused public OCT database. The obvious gap in applications EF recognition motivated us design publish a comprehensive...
As a non-invasive optical imaging technique, coherence tomography (OCT) has proven promising for automatic fingerprint recognition system (AFRS) applications. Diverse approaches have been proposed OCT-based presentation attack detection (PAD). However, considering the complexity and variety of PA samples, it is extremely challenging to increase generalization ability with limited dataset. To solve challenge, this paper presents novel supervised learning-based PAD method, denoted as internal...
Raman spectroscopy is a powerful tool for identifying substances, yet accurately analyzing mixtures remains challenging due to overlapping spectra. This study aimed develop deep learning-based framework improve the identification of components in using spectroscopy. We propose three-branch feature fusion network that leverages spectral pairwise comparison and multi-head self-attention mechanism capture both local global features. To address limited data availability, traditional augmentation...