Historically, the research group has focused primarily on research in the areas of biometrics, sensor technologies, and security applications. The current STRaDe research team covers three interrelated areas. Given the importance of the practical application of our results, we actively collaborate with specialists who can put them into practice.
Research areas
Biometrics and Forensic Data Analysis
- Finger Vein Recognition
- Development of recognition algorithms for forensic applications
- Liveness Detection and Spoofing of Biometric Characteristics
About this area and its results
The development of neural networks in recent years has had a significant impact on biometric research, opening up new possibilities in the field. So far, we have focused on a wide range of biometric topics, from the development of proprietary recognition algorithms to liveness detection and sensor reliability assessment. Forensic topics include scene context analysis and the development of new methods for trace comparison and digitization.
Selected results
- Head Poses and Grimaces: Challenges for automated face identification algorithms?, publication, 2024,Detail
- Facial Image Reconstruction and its Influence to Face Recognition, publication, 2023,Detail
- FaceAIKit: Library for face detection and recognition, software, 2023, Detail
- State-of-the-Art in 3D Finger Vein Recognition and a Novel Mirror-Based 3D Finger Vein Imaging Device, publication, 2022,Detail
- On the fly biometric identification system using hand-geometry, publication, 2021,Detail
Sensors and Edge Computing
- Sensor Device for Laser Beam Detection in Open Space
- Edge Computing Technology
About this area and its results
With the development of neural network-based algorithms, tools for fast inference (computation) of neural networks have emerged, opening up new possibilities not only for biometric devices.
Selected results
- Detekce laserového svazku v otevřeném prostoru, software, 2023, Detail
- Multimodální snímač prostorové struktury krevního řečiště prstu, užitný vzor, 2023, Detail
- Zařízení pro biometrickou identifikaci pomocí otisků prstů a/nebo vlastností rukou a metody pro biometrickou identifikaci pomocí těchto charakteristik, patent, 2024, Detail
- Phenobottle 2: Algae cultivator for distributed optimization with temperature control and spectral density measurement, prototyp, 2022, Detail
Medical Data Analysis
- Analysis of Biomedical Imaging Data
- Reconstruction of Tracheal Sections for Monitoring Stenosis Progression
About this area and its results
Medical data processing is a discipline that combines detection, classification, and segmentation, where the integration of these methods with text models plays a key role. The goal of the research is to improve the diagnostic capabilities for various diseases.
Selected results
Technologies
As part of our research and applied development activities, we utilize advanced technologies that bridge experimental research with practical deployment. Some of these have become the foundation of our research projects, prototypes, and industrial applications. Examples include the following technologies.
For Research
- Advanced Neural Architectures: Vision Transformer (ViT), Swin Transformer, ArcFace, AdaFace
- Generative and Forensic Models:Diffusion and GAN networks for image synthesis and manipulation detection
- Biomedical Data Analysis:MONAI, nnU-Net, 3D Slicer – integration of segmentation and reconstruction methods
- Multimodal Learning and Sensor Fusion:combination of visual, textual, and physiological data
For Deployment
- Edge AI Platforms:NVIDIA Jetson, Intel OpenVINO, TensorRT – model optimization for real-world operation
- Distributed Processing and Orchestration: Docker, Podman
- Research Data Infrastructure:MLflow, Weights & Biases – experiment tracking, versioning, and reproducibility
- Forensic and Biometric Standards: ISO/IEC 19794, NIST NBIS, BioLab toolkit
Collaboration
For students
Join our research through a bachelor’s or master’s thesis, or a project internship. Explore available topics in biometrics, sensors and medical data analysis.
View thesis topicsFor companies and partners
We collaborate on applied research and technology development to address practical challenges. Email us about the problem you need to solve, and we will discuss opportunities to work together.
strade@fit.vut.cz