A5027730478- Biometric Identification and Security
- Face recognition and analysis
- Advanced Vision and Imaging
- Video Surveillance and Tracking Methods
- Generative Adversarial Networks and Image Synthesis
- Face and Expression Recognition
- Retinal Imaging and Analysis
- 3D Surveying and Cultural Heritage
- Advanced Image and Video Retrieval Techniques
- Handwritten Text Recognition Techniques
- Digital Media Forensic Detection
- User Authentication and Security Systems
- Gait Recognition and Analysis
- Advanced Neural Network Applications
- Robotics and Sensor-Based Localization
- Reconstructive Facial Surgery Techniques
- Vehicle License Plate Recognition
- Color Science and Applications
- Image Retrieval and Classification Techniques
- Human Pose and Action Recognition
- Optical measurement and interference techniques
- 3D Shape Modeling and Analysis
- Industrial Vision Systems and Defect Detection
- Forensic Fingerprint Detection Methods
- Visual Attention and Saliency Detection
University of Ljubljana
2016-2025
College for Management in Tourism and Informatics in Virovitica
2021
The University of Texas at Dallas
2020
University of Maribor
2020
XLAB (Slovenia)
2020
Griffith University
2018
Universidad de Navarra
2005-2007
Centro de Estudios e Investigaciones Técnicas de Gipuzkoa
2005-2007
Computer vision is one of many areas that wants to understand the process human functionality and copy with intention complement life intelligent machines. For better human-computer interaction it necessary for machine see people. This can be achieved by employing face detection algorithms, like used in installation "15 Seconds Fame". unites modern art technology. Its algorithm based on skin colour detection. One problems this similar algorithms have deal sensitivity illumination conditions...
Biometric recognition technology has made significant advances over the last decade and is now used across a number of services applications. However, this widespread deployment also resulted in privacy concerns evolving societal expectations about appropriate use technology. For example, ability to automatically extract age, gender, race, health cues from biometric data heightened leakage. Face technology, particular, been spotlight, seen by many as posing considerable risk personal...
The new Coronavirus disease (COVID-19) has seriously affected the world. By end of November 2020, global number coronavirus cases had already exceeded 60 million and deaths 1,410,378 according to information from World Health Organization (WHO). To limit spread disease, mandatory face-mask rules are now becoming common in public settings around Additionally, many service providers require customers wear face-masks accordance with predefined (e.g., covering both mouth nose) when using...
Image-based virtual try-on techniques have shown great promise for enhancing the user-experience and improving customer satisfaction on fashion-oriented e-commerce platforms. However, existing are currently still limited in quality of results they able to produce from input images diverse characteristics. In this work, we propose a Context-Driven Virtual Try-On Network (C-VTON) that addresses these limitations convincingly transfers selected clothing items target subjects even under...
Image and video data are today being shared between government entities other relevant stakeholders on a regular basis require careful handling of the personal information contained therein. A popular approach to ensure privacy protection in such is use deidentification techniques, which aim at concealing identity individuals imagery while still preserving certain aspects after deidentification. In this work, we propose novel towards face deidentification, called k-Same-Net, combines recent...
In this paper we present the results of Unconstrained Ear Recognition Challenge (UERC), a group benchmarking effort centered around problem person recognition from ear images captured in uncontrolled conditions. The goal challenge was to assess performance existing techniques on challenging large-scale dataset and identify open problems that need be addressed future. Five groups three continents participated contributed six for evaluation, while multiple baselines were made available by UERC...
Identity recognition from ear images is an active field of research within the biometric community. The ability to capture a distance and in covert manner makes technology appealing choice for surveillance security applications as well related application domains. In contrast other modalities, where large datasets captured uncontrolled settings are readily available, still limited size mostly laboratory-like quality. As consequence, has not benefited yet advances deep learning...
Iris segmentation is an important research topic that received significant attention from the community over years. Traditional iris techniques have typically been focused on hand-crafted procedures that, nonetheless, achieved remarkable performance even with images captured in difficult settings. With success of deep-learning models, researchers are increasingly looking towards convolutional neural networks (CNNs) to further improve accuracy existing and several CNN-based already presented...
Segmentation techniques for ocular biometrics typically focus on finding a single eye region in the input image at time. Only limited work has been done multi-class segmentation despite number of obvious advantages. In this paper we address gap and present deep model build around SegNet architecture. We train small dataset (of 120 samples) images observe it to generalize well unseen ensure highly accurate results. evaluate Multi-Angle Sclera Database (MASD) describe comprehensive experiments...
Object detection and segmentation represents the basis for many tasks in computer machine vision. In biometric recognition systems of region-of-interest (ROI) is one most crucial steps processing pipeline, significantly impacting performance entire system. Existing approaches to ear detection, are commonly susceptible presence severe occlusions, accessories or variable illumination conditions often deteriorate their if applied on images captured unconstrained settings. To address these...
Convolutional Neural Networks (CNNs) are today the de-facto standard for extracting compact and discriminative face representations (templates) from images in automatic recognition systems. Due to characteristics of CNN models, generated typically encode a multitude information ranging identity soft-biometric attributes, such as age, gender or ethnicity. However, since these were computed purpose only, contained templates represents serious privacy risk. To mitigate this problem, we present...
This paper presents a summary of the Competition on Face Morphing Attack Detection Based Privacy-aware Synthetic Training Data (SYN-MAD) held at 2022 In-ternational Joint Conference Biometrics (IJCB 2022). The competition attracted total 12 participating teams, both from academia and industry present in 11 differ-ent countries. In end, seven valid submissions were submitted by teams evaluated organizers. was to at-tract solutions that deal with detecting face morphing at-tacks while...
Ear recognition is a contactless and unobtrusive biometric technique with applications across various domains. However, deploying high-performing ear models on resource-constrained devices challenging, limiting their applicability widespread adoption. This paper introduces EdgeEar, lightweight model based proposed hybrid CNN-transformer architecture to solve this problem. By incorporating low-rank approximations into specific linear layers, EdgeEar reduces its parameter count by factor of 50...
The large majority of modern software solutions intended for fingermark processing in a forensic context is heavily dependant on the correct image scaling. Fingermark images captured with digital cameras at crime scene require use physical rulers or labels. While resolution can be calibrated manually by examiner lab, we propose an automated approach, which could integrated directly into existing identification systems and would eliminate need human intervention. Our approach consists CNN...
Computer vision and biometrics are increasingly important in many AI-driven applications, yet teaching these fields poses challenges balancing theory hands-on practice. This paper presents a structured approach implemented for the technical skills course at Faculty of Information Science, University Ljubljana, designed Science students. The integrates guided Jupyter Notebook exercises while allowing students to complete coding tasks leaning on AI assistance. In-person presentations...
With the rising computational and memory cost of deep neural networks there is more effort to reduce size these models, especially when their deployment on resource constrained devices goal. New methods compressing are being constantly developed with goal minimizing drop in accuracy. In this paper we focus pruning techniques as a way compression. We present comparison different criteria analyze loss accuracy for case simple non-iterative procedure. provide between cases applied architectures...
This paper summarises the results of Sclera Segmentation and Eye Recognition Benchmarking Competition (SSERBC 2017). It was organised in context International Joint Conference on Biometrics (IJCB The aim this competition to record recent developments sclera segmentation eye recognition visible spectrum (using iris, peri-ocular, their fusion), also gain attention researchers subject. In regard, we have used Multi-Angle Dataset (MASD version 1). is comprised of2624 images taken from both eyes...
Modern face recognition (FR) models excel in constrained scenarios, but often suffer from decreased performance when deployed unconstrained (real-world) environments due to uncertainties surrounding the quality of captured facial data. Face image assessment (FIQA) techniques aim mitigate these degradations by providing FR with sample-quality predictions that can be used reject low-quality samples and reduce false match errors. However, despite steady improvements, ensuring reliable estimates...