A5068649110- Biometric Identification and Security
- Digital Media Forensic Detection
- Sports Performance and Training
- Face recognition and analysis
- Cell Image Analysis Techniques
- Handwritten Text Recognition Techniques
- Sports Analytics and Performance
- Image Retrieval and Classification Techniques
- Video Analysis and Summarization
- Advanced Image and Video Retrieval Techniques
- Heart Rate Variability and Autonomic Control
- Sports injuries and prevention
- Advanced Steganography and Watermarking Techniques
- Advanced Neural Network Applications
- Anomaly Detection Techniques and Applications
- Cardiovascular and exercise physiology
- Microplastics and Plastic Pollution
- User Authentication and Security Systems
- Human Pose and Action Recognition
- Natural Language Processing Techniques
- Algorithms and Data Compression
- Image Processing and 3D Reconstruction
- Forensic and Genetic Research
- Generative Adversarial Networks and Image Synthesis
- Advanced Chemical Sensor Technologies
Universidade Estadual de Campinas (UNICAMP)
2013-2024
Brazilian Center for Research in Energy and Materials
2022-2024
Brazilian Synchrotron Light Laboratory
2022-2024
ORCID
2020-2021
Hospital de Clínicas da Unicamp
2019
University of Notre Dame
2017
Biometrics systems have significantly improved person identification and authentication, playing an important role in personal, national, global security. However, these might be deceived (or spoofed) and, despite the recent advances spoofing detection, current solutions often rely on domain knowledge, specific biometric reading systems, attack types. We assume a very limited knowledge about at sensor to derive outstanding detection for iris, face, fingerprint modalities based two deep...
Despite important recent advances, the vulnerability of biometric systems to spoofing attacks is still an open problem. Spoof occur when impostor users present synthetic samples a valid user system seeking deceive it. Considering case face biometrics, attack consists in presenting fake sample (e.g., photograph, digital video, or even 3D mask) acquisition sensor with facial information user. In this paper, we introduce low cost and software-based method for detecting attempts recognition...
Spoofing attacks or impersonation can be easily accomplished in a facial biometric system wherein users without access privileges attempt to authenticate themselves as valid users, which an impostor needs only photograph video with information of legitimate user. Even recent advances biometrics, forensics and security, vulnerability systems against spoofing is still open problem. though several methods have been proposed for photo-based attack detection, performed videos vastly overlooked,...
The study of brain connectivity has been a cornerstone in understanding the complexities neurological and psychiatric disorders. It provided invaluable insights into functional architecture how it is perturbed However, persistent challenge achieving proper spatial resolution, developing computational algorithms to address biological questions at multi-cellular level, scale often referred as mesoscale. Historically, neuroimaging studies have predominantly focused on macroscale, providing...
As a crucial security problem, anti-spoofing in biometrics, and particularly for the face modality, has achieved great progress recent years. Still, new threats arrive inform of better, more realistic sophisticated spoofing attacks. The objective 2nd Competition on Counter Measures to 2D Face Spoofing Attacks is challenge researchers create counter measures effectively detecting variety submitted propositions are evaluated Replay-Attack database results presented this paper.
Recent advances on biometrics, information forensics, and security have improved the accuracy of biometric systems, mainly those based facial information. However, an ever-growing challenge is vulnerability such systems to impostor attacks, in which users without access privileges try authenticate themselves as valid users. In this work, we present a solution video-based face spoofing systems. Such type attack characterized by presenting video real user system. To best our knowledge, first...
The adoption of large-scale iris recognition systems around the world has brought to light importance detecting presentation attack images (textured contact lenses and printouts). This work presents a new approach in Presentation Attack Detection (PAD), by exploring combinations Convolutional Neural Networks (CNNs) transformed input spaces through binarized statistical image features (BSIF). Our method combines lightweight CNNs classify multiple BSIF views image. Following explorations on...
Prior art has shown it is possible to estimate, through image processing and computer vision techniques, the types parameters of transformations that have been applied content individual images obtain new images. Given a large corpus query image, an interesting further step retrieve set original whose present in as well detailed sequences yield given This problem recently received name provenance analysis. In these times public media manipulation ( <italic...
Pollution in the form of litter natural environment is one great challenges our times. Automated detection can help assess waste occurrences environment. Different machine learning solutions have been explored to develop tools, thereby supporting research, citizen science, and volunteer clean-up initiatives. However, best knowledge, no work has investigated performance state-of-the-art deep object approaches context detection. In particular, studies focused on assessment those methods aiming...
Presentation attack detection is a challenging problem that aims at exposing an impostor user seeking to deceive the authentication system. In facial biometrics systems, this kind of performed using photograph, video, or 3D mask containing biometric information genuine identity. paper, we propose novel approach detecting face presentation attacks based on intrinsic properties scene such as albedo, depth, and reflectance surfaces, which were recovered through shape-from-shading (SfS)...
The aim of this study was to identify the most relevant variables characterise performance level teams through Men's World Championships (2007–2019). Forty-seven attributes from match-related statistics and characteristics players were analysed in 168 participant teams. Descriptive discriminant analysis classified correctly 69.6% cases selected height players, 9-m efficiency, international matches disputed, wing blocked shots, 7-m goalkeeper efficiency 2-min suspensions which indicators....
Kicking is a fundamental skill in soccer that often contributes to match outcomes. Lower limb movement features (e.g., joint position and velocity) are determinants of kick performance. However, obtaining kicking kinematics under field conditions generally requires time-consuming manual tracking. The current study aimed compare contemporary markerless automatic motion estimation algorithm (OpenPose) with digitisation (DVIDEOW software) on-field kinematic parameters. An experimental dataset...
Cardiopulmonary exercise testing (CPET) is a non-invasive approach to measure the maximum oxygen uptake ([Formula: see text]), which an index assess cardiovascular fitness (CF). However, CPET not available all populations and cannot be obtained continuously. Thus, wearable sensors are associated with machine learning (ML) algorithms investigate CF. Therefore, this study aimed predict CF by using ML data technologies. For purpose, 43 volunteers different levels of aerobic power, who wore...
The use of marker-less methods to automatically obtain kinematics movement is expanding but validity high-velocity tasks such as cycling with the presence bicycle on field view needed when standard video footage obtained. purpose this study was assess if pre-trained neural networks are valid for calculations lower limb joint during cycling. Motion twenty-six cyclists pedalling a cycle trainer captured by camera capturing frames from sagittal plane whilst reflective markers were attached...
This article addresses the pixelwise classification problem based on temporal profiles, which are encoded in 2-D representations recurrence plots, Gramian angular/ difference fields, and Markov transition field. We propose a multirepresentational fusion scheme that exploits complementary view provided by those time series different data-driven feature extractors classifiers. validate our ensemble related to of eucalyptus plantations remote sensing images. Achieved results demonstrate...
Sports sciences are increasingly data-intensive nowadays since computational tools can extract information from large amounts of data and derive insights athlete performances during the competition. This paper addresses a performance prediction problem in soccer, popular collective sport modality played by two teams competing against each other same field. In soccer game, score points placing ball into opponent’s goal winner is team with highest count goals. Retaining possession one key to...
X-ray computed microtomography (μCT) is an innovative and nondestructive versatile technique that has been used extensively to investigate bio-based systems in multiple application areas. Emerging progress this field brought countless studies using μCT characterization, revealing three-dimensional (3D) material structures quantifying features such as defects, pores, secondary phases, filler dispersions, internal interfaces. Recently, x-ray tomography (CT) beamlines coupled synchrotron light...
This study investigated the 30-days altitude training (2500 m, LHTH-live and high) on hematological responses aerobic-anaerobic performances parameters of high-level Paralympic athletes. Aerobic capacity was assessed by 3000 m run, anaerobic variables (velocity, force mechanical power) a maximal 30-s semi-tethered running test (AO30). These assessments were carried out at low before (PRE) after LHTH (5-6 15-16 days, POST1 POST2, respectively). During LHTH, analyzes performed days 1, 12, 20...
Deriving relationships between images and tracing back their history of modifications are at the core Multimedia Phylogeny solutions, which aim to combat misinformation through doctored visual media. Nonetheless, most recent image phylogeny solutions cannot properly address cases forged composite with multiple donors, an area known as parenting (MPP). This paper presents a preliminary undirected graph construction solution for MPP, without any strict assumptions. The algorithm is underpinned...
Scene text detection has become an important field in the computer vision area due to increasing number of applications. This is a very challenging problem as textual elements are commonly found "noisy" and complex natural scenes. Another issue refers presence texts encoded into different languages within same image. State-of-the-art solutions rely on use deep neural network approaches or even ensembles them. However, such associated with "heavy" models, which computationally expensive terms...
Although several studies have focused on the adaptations provided by inspiratory muscle (IM) training physical demands, warm-up or pre-activation (PA) of these muscles alone appears to generate positive effects physiological responses and performance. This study aimed understand (IMPA) high-intensity running passive recovery, as applied active subjects. In an original innovative investigation impacts IMPA running, we proposed identification interactions among characteristics, oxygenation in...
Abstract This study investigated the effects of inspiratory muscle pre-activation (IM PA ) on interactions among technical-tactical, physical, physiological, and psychophysiological parameters in a simulated judo match, based centrality metrics by complex network model. Ten male athletes performed 4 experimental sessions. Firstly, anthropometric measurements, maximal pressure (MIP) global strenght muscles were determined. In following days, all four-minute video-recorded matches, under three...
The present study aimed to assess the use of technical-tactical variables and machine learning (ML) classifiers in automatic classification passing difficulty (DP) level soccer matches illustrate model with best performance distinguish players. We compared eight ML according their accuracy classifying events using 35 based on spatiotemporal data. Support Vector Machine (SVM) algorithm achieved a balanced 0.70 ± 0.04%, considering multi-class classification. Next, we best-performing...