A5044741265- Video Surveillance and Tracking Methods
- Advanced Image and Video Retrieval Techniques
- Human Pose and Action Recognition
- Face recognition and analysis
- Face and Expression Recognition
- Anomaly Detection Techniques and Applications
- Advanced Neural Network Applications
- Biometric Identification and Security
- Gait Recognition and Analysis
- Advanced Vision and Imaging
- Remote-Sensing Image Classification
- Image Retrieval and Classification Techniques
- Vehicle License Plate Recognition
- Handwritten Text Recognition Techniques
- Domain Adaptation and Few-Shot Learning
- Social Work Education and Practice
- Context-Aware Activity Recognition Systems
- Video Analysis and Summarization
- Robotics and Sensor-Based Localization
- Digital Media Forensic Detection
- Medical Image Segmentation Techniques
- Remote Sensing and Land Use
- Multimodal Machine Learning Applications
- Mathematical Dynamics and Fractals
- Cardiac Arrhythmias and Treatments
Universidade Federal de Minas Gerais
2015-2024
Illinois Institute of Technology
2019
Laboratoire d'Informatique de Paris-Nord
2016
Centro Universitário de Belo Horizonte
2012-2015
Universidade Estadual de Campinas (UNICAMP)
2011-2012
Hospital de Clínicas da Unicamp
2012
University of Maryland, College Park
2006-2010
Williams (United States)
2010
Universidade Federal do Paraná
2001-2005
Fairfield University
2003
An electrocardiogram (ECG) measures the electric activity of heart and has been widely used for detecting diseases due to its simplicity non-invasive nature. By analyzing electrical signal each heartbeat, i.e., combination action impulse waveforms produced by different specialized cardiac tissues found in heart, it is possible detect some abnormalities. In last decades, several works were developed produce automatic ECG-based heartbeat classification methods. this work, we survey current...
Significant research has been devoted to detecting people in images and videos. In this paper we describe a human detection method that augments widely used edge-based features with texture color information, providing us much richer descriptor set. This augmentation results an extremely high-dimensional feature space (more than 170,000 dimensions). such spaces, classical machine learning algorithms as SVMs are nearly intractable respect training. Furthermore, the number of training samples...
Appearance information is essential for applications such as tracking and people recognition. One of the main problems using appearance-based discriminative models ambiguities among classes when number persons being considered increases. To reduce amount ambiguity, we propose use a rich set feature descriptors based on color, textures edges. Another issue regarding appearance modeling limited training samples available each appearance. The are created powerful statistical tool called partial...
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...
Automatic License Plate Recognition (ALPR) has been a frequent topic of research due to many practical applications. However, the current solutions are still not robust in real-world situations, commonly depending on constraints. This paper presents and efficient ALPR system based state-of-the-art YOLO object detector. The Convolutional Neural Networks (CNNs) trained finetuned for each stage so that they under different conditions (e.g., variations camera, lighting, background). Specially...
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...
Due to the availability of large-scale skeleton datasets, 3D human action recognition has recently called attention computer vision community. Many works have focused on encoding data as image representations based spatial structure joints, in which temporal dynamics sequence is encoded variations columns and each frame represented rows a matrix. To further improve such representations, we introduce novel representation be used input Convolutional Neural Networks (CNNs), named SkeleMotion....
This paper presents an approach for detecting anomalous events in videos with crowds. The main goal is to recognize patterns that might lead event. An event be characterized by the deviation from normal or usual, but not necessarily undesirable manner, e.g., just different a suspicious surveillance point of view. One challenges such difficulty create models due their unpredictability and dependency on context scene. Based these challenges, we present model uses general concepts, as...
In the last years, computer vision research community has studied on how to model temporal dynamics in videos employ 3D human action recognition. To that end, two main baseline approaches have been researched: (i) Recurrent Neural Networks (RNNs) with Long-Short Term Memory (LSTM); and (ii) skeleton image representations used as input a Convolutional Network (CNN). Although RNN present excellent results, such methods lack ability efficiently learn spatial relations between joints. On other...
Abstract This paper presents an efficient and layout‐independent Automatic License Plate Recognition (ALPR) system based on the state‐of‐the‐art you only look once (YOLO) object detector that contains a unified approach for license plate (LP) detection layout classification to improve recognition results using post‐processing rules. The is conceived by evaluating optimizing different models, aiming at achieving best speed/accuracy trade‐off each stage. networks are trained images from...
Chronic congestive heart failure not controlled by conventional therapy was treated with intravenous amrinone, a new non-glycosidic, non-catecholamine cardiotonic agent. Eight patients New York Heart Association functional class III-IV symptoms were hemodynamically monitored. At peak effect, cardiac index (CI) increased from 1.84 +/- 0.32 to 2.74 0.44 l/min/m2 (mean SD) (p less than 0.001) and left ventricular filling pressure (LVFP) decreased 25.8 6.2 19.5 6.8 mm Hg 0.05), while rate mean...
Spoofing identities using photographs is one of the most common techniques to attack 2-D face recognition systems. There seems exist no comparative studies different same protocols and data. The motivation behind this competition compare performance state-of-the-art algorithms on database a unique evaluation method. Six teams from universities around world have participated in contest. Use or multiple motion, texture analysis liveness detection appears be trend competition. Most are able...
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,...
Since the beginning of new millennium, electrocardiogram (ECG) has been studied as a biometric trait for security systems and other applications. Recently, with devices such smartphones tablets, acquisition ECG signal in off-the-person category made this suitable real scenarios. In paper, we introduce usage deep learning techniques, specifically convolutional networks, extracting useful representation heart biometrics recognition. Particularly, investigate feature representations through two...
Personal identity verification based on biometrics has received increasing attention since it allows reliable authentication through intrinsic characteristics, such as face, voice, iris, fingerprint, and gait. Particularly, face recognition techniques have been used in a number of applications, security surveillance, access control, crime solving, law enforcement, among others. To strengthen the results verification, biometric systems must be robust against spoofing attempts with photographs...
With the goal of matching unknown faces against a gallery known people, face identification task has been studied for several decades. There are very accurate techniques to perform in controlled environments, particularly when large numbers samples available each face. However, under uncontrolled environments or with lack training data is still an unsolved problem. We employ and rich set feature descriptors (with more than 70,000 descriptors) using partial least squares multichannel...
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.
Although visible face recognition has been an active area of research for several decades, cross-modal only explored by the biometrics community relatively recently. Thermal-to-visible is one most difficult challenges, because difference in phenomenology between thermal and imaging modalities. We address problem using a partial least squares (PLS) regression-based approach consisting preprocessing, feature extraction, PLS model building. The preprocessing extraction stages are designed to...
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...
Human activity recognition based on wearable sensor data has been an attractive research topic due to its application in areas such as healthcare and smart environments. In this context, many works have presented remarkable results using accelerometer, gyroscope magnetometer represent the activities categories. However, current studies do not consider important issues that lead skewed results, making it hard assess quality of sensor-based human preventing a direct comparison previous works....
Automatic License Plate Recognition (ALPR) has been the focus of many researches in past years. In general, ALPR is divided into following problems: detection on-track vehicles, license plates detection, segmention plate characters and optical character recognition (OCR). Even though commercial solutions are available for controlled acquisition conditions, e.g., entrance a parking lot, still an open problem when dealing with data acquired from uncontrolled environments, such as roads...
Semantic segmentation requires methods capable of learning high-level features while dealing with large volume data. Toward such goal, convolutional networks can learn specific and adaptable based on the However, these are not processing a whole remote sensing image, given its huge size. To overcome limitation, image is processed using fixed size patches. The definition input patch usually performed empirically (evaluating several sizes) or imposed (by network constraint). Both strategies...