Cancer is the second leading cause of death in US after cardiovascular disease. Image-based computer-aided diagnosis can assist physicians to efficiently diagnose cancers in early stages. Existing computer-aided algorithms use... more
Cancer is the second leading cause of death in US after cardiovascular disease. Image-based computer-aided diagnosis can assist physicians to efficiently diagnose cancers in early stages. Existing computer-aided algorithms use hand-crafted features such as wavelet coefficients, co-occurrence matrix features, and recently, histogram of shearlet coefficients for classification of cancerous tissues and cells in images. These hand-crafted features often lack generalizability since every cancerous tissue and cell has a specific texture, structure, and shape. An alternative approach is to use convolutional neural networks (CNNs) to learn the most appropriate feature abstractions directly from the data and handle the limitations of hand-crafted features. A framework for breast cancer detection and prostate Gleason grading using CNN trained on images along with the magnitude and phase of shearlet coefficients is presented. Particularly, we apply shearlet transform on images and extract the ...
Research Interests:
Subthalamic nucleus (STN) local field potentials (LFP) are neural signals that have been shown to reveal motor and language behavior, as well as pathological parkinsonian states. We use a research-grade implantable neurostimulator (INS)... more
Subthalamic nucleus (STN) local field potentials (LFP) are neural signals that have been shown to reveal motor and language behavior, as well as pathological parkinsonian states. We use a research-grade implantable neurostimulator (INS) with data collection capabilities to record STN-LFP outside the operating room to determine the reliability of the signals over time and assess their dynamics with respect to behavior and dopaminergic medication. Seven subjects were implanted with the recording augmented deep brain stimulation (DBS) system, and bilateral STN-LFP recordings were collected in the clinic over twelve months. Subjects were cued to perform voluntary motor and language behaviors in on and off medication states. The STN-LFP recorded with the INS demonstrated behavior-modulated desynchronization of beta frequency (13-30 Hz) and synchronization of low gamma frequency (35-70 Hz) oscillations. Dopaminergic medication did not diminish the relative beta frequency oscillatory desyn...
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests: Engineering, Computer Science, Computational Complexity, Graph Theory, Face Recognition, and 16 moreMesh generation, Biometrics, Dempster-Shafer Analysis, Face Modeling, Image fusion, Wavelet Transforms, Data acquisition, Facial Features, Feature Extraction, Active Shape Model, Hausdorff Distance, Range Image, Delaunay Triangulation, Gaussian Curvature, Information Forensics, and Gabor Wavelets
Research Interests:
Research Interests:
Research Interests:
Abstract We present an automatic disparity-based approach for 3D face modeling, from two frontal and one profile view stereo images, for 3D face recognition applications. Once the images are captured, the algorithm starts by extracting... more
Abstract We present an automatic disparity-based approach for 3D face modeling, from two frontal and one profile view stereo images, for 3D face recognition applications. Once the images are captured, the algorithm starts by extracting selected 2D facial features from ...
Research Interests:
Research Interests:
Research Interests:
... 25, no. 6, pp. 583 594, 1992. [20] Arun A. Ross, Karthik Nandakumar, and Anil K. Jain, Hand-book of Multibiometrics (International Series on Biometrics), Springer-Verlag New York, Inc., Secaucus, NJ, USA, 2006. [21] A-Nansser Ansari,... more
... 25, no. 6, pp. 583 594, 1992. [20] Arun A. Ross, Karthik Nandakumar, and Anil K. Jain, Hand-book of Multibiometrics (International Series on Biometrics), Springer-Verlag New York, Inc., Secaucus, NJ, USA, 2006. [21] A-Nansser Ansari, M. Abdel-Mottaleb, and Mohammad H ...
Research Interests:
Research Interests:
Research Interests: Psychology, Computer Vision, Face Recognition, Image Analysis, Support Vector Machines, and 12 moreFacial expression, Behavioral Science, Face Detection, Face to Face, Computer Vision and Pattern Recognition, Regression Analysis, Shape, Support vector machine, Gold Standard, High Dimensionality, Facial Action Coding System, and Action Unit
ABSTRACT Automatic facial expression analysis has received great attention in different applications over the last two decades. Facial Action Coding System (FACS), which describes all possible facial expressions based on a set of facial... more
ABSTRACT Automatic facial expression analysis has received great attention in different applications over the last two decades. Facial Action Coding System (FACS), which describes all possible facial expressions based on a set of facial muscle movements called Action Unit (AU), has been used extensively to model and analyze facial expressions. FACS describes methods for coding the intensity of AUs, and AU intensity measurement is important in some studies in behavioral science and developmental psychology. However, in majority of the existing studies in the area of facial expression recognition, the focus has been on basic expression recognition or facial action unit detection. There are very few investigations on measuring the intensity of spontaneous facial actions. In addition, the few studies on AU intensity recognition usually try to measure the intensity of facial actions statically and individually, ignoring the dependencies among multilevel AU intensities as well as the temporal information. However, these spatiotemporal interactions among facial actions are crucial for understanding and analyzing spontaneous facial expressions, since these coherent, coordinated, and synchronized interactions are that produce a meaningful facial display. In this paper, we propose a framework based on Dynamic Bayesian Network (DBN) to systematically model the dynamic and semantic relationships among multilevel AU intensities. Given the extracted image observations, the AU intensity recognition is accomplished through probabilistic inference by systematically integrating the image observations with the proposed DBN model. Experiments on Denver Intensity of Spontaneous Facial Action (DISFA) database demonstrate the superiority of our method over single image-driven methods in AU intensity measurement.
Research Interests:
Research Interests:
ABSTRACT Automatic recognition of facial expressions is an interesting and challenging research topic in the field of pattern recognition due to applications such as human-machine interface design and developmental psychology. Designing... more
ABSTRACT Automatic recognition of facial expressions is an interesting and challenging research topic in the field of pattern recognition due to applications such as human-machine interface design and developmental psychology. Designing classifiers for facial expression recognition with high reliability is a vital step in this research. This paper presents a novel framework for person-independent expression recognition by combining multiple types of facial features via multiple kernel learning (MKL) in multiclass support vector machines (SVM). Existing MKL-based approaches jointly learn the same kernel weights with \(l_{1}\) -norm constraint for all binary classifiers, whereas our framework learns one kernel weight vector per binary classifier in the multiclass-SVM with \(l_{p}\) -norm constraints \((p \ge 1)\) , which considers both sparse and non-sparse kernel combinations within MKL. We studied the effect of \(l_{p}\) -norm MKL algorithm for learning the kernel weights and empirically evaluated the recognition results of six basic facial expressions and neutral faces with respect to the value of “ \(p\) ”. In our experiments, we combined two popular facial feature representations, histogram of oriented gradient and local binary pattern histogram, with two kernel functions, the heavy-tailed radial basis function and the polynomial function. Our experimental results on the CK \(+\) , MMI and GEMEP-FERA face databases as well as our theoretical justification show that this framework outperforms the state-of-the-art methods and the SimpleMKL-based multiclass-SVM for facial expression recognition.
Research Interests:
Automatic measurement of spontaneous facial action units (AUs) defined by the facial action coding system (FACS) is a challenging problem. The recent FACS user manual defines 33 AUs to describe different facial activities and expressions.... more
Automatic measurement of spontaneous facial action units (AUs) defined by the facial action coding system (FACS) is a challenging problem. The recent FACS user manual defines 33 AUs to describe different facial activities and expressions. In spontaneous facial expressions, a subset of AUs are often occurred or activated at a time. Given this fact that AUs occurred sparsely over time, we propose a novel method to detect the absence and presence of AUs and estimate their intensity levels via sparse representation (SR). We use the robust principal component analysis to decompose expression from facial identity and then estimate the intensity of multiple AUs jointly using a regression model formulated based on dictionary learning and SR. Our experiments on Denver intensity of spontaneous facial action and UNBC-McMaster shoulder pain expression archive databases show that our method is a promising approach for measurement of spontaneous facial AUs.
Research Interests:
This paper presents an approach for measuring and monitoring human body joint angles using inertial measurement unit (IMU) sensors. This type of monitoring is beneficial for therapists and physicians because it facilitates remote... more
This paper presents an approach for measuring and monitoring human body joint angles using inertial measurement unit (IMU) sensors. This type of monitoring is beneficial for therapists and physicians because it facilitates remote assessment of patient activities. In our approach, two IMUs are mounted on the upper leg and the lower leg to measure the Euler angles of each segment. The Euler angles are sent via Bluetooth protocols to a pc for calculating the knee joint angle. In our experiments, we utilized a motion capture system to accurately measure the knee joint angle and used this as the ground truth to assess the accuracy of the IMU system. The range of average error of the system across a variety of motion trials was 0.08 to 3.06 degrees. In summary, the accuracy of the IMU measurement system currently outperforms existing wearable systems such as conductive fiber optic sensors and flex-sensors.
Research Interests: Algorithms, Biomechanics, Sensors, Computers, Humans, and 18 moreTransducers, Movement, Human Body, Fiber optic sensor, Patient Monitoring, Motion, Acceleration, Motion Capture, Time Factors, Reproducibility of Results, Ambulatory BP monitoring, Ground Truth, Fiber Optic Technology, Knee Joint, Equipment Design, Inertial Measurement Unit, Biomechanical Phenomena, and Measurement System
With a growing number of low-income patients developing Congestive Heart Failure in urban Denver, accessible and affordable solutions are needed to provide home management options. A multidisciplinary team evaluated currently available... more
With a growing number of low-income patients developing Congestive Heart Failure in urban Denver, accessible and affordable solutions are needed to provide home management options. A multidisciplinary team evaluated currently available options for telemonitoring and developed a solution for an initial pilot study. This system is currently used in the Denver Metro area (Colorado) for 44 CHF patients. Preliminary results show this approach is effective and has reduced the patients' average length of stay at the hospital compared to historical data and control patients who do not use a remote monitoring system.
Research Interests: Cardiology, Chronic illness, Heart Failure, Humans, Congestive Heart Failure, and 13 morePatient Monitoring, Length of Stay, Health monitoring, Historical Data, Colorado, Pilot study, Cardiovascular system, Remote Monitoring, Urban Population, Multidisciplinary Teams, Low Income, Pilot Projects, and Home Monitoring
Research Interests:
Research Interests:
... Mohammad H. Mahoor and Mohamed Abdel-Mottaleb Department of Electrical and Computer Engineering,University of Miami 1251 Memorial Drive, Coral Gables, FL ... Therefore, we need to find the optimum parameters of the ASM that best fit... more
... Mohammad H. Mahoor and Mohamed Abdel-Mottaleb Department of Electrical and Computer Engineering,University of Miami 1251 Memorial Drive, Coral Gables, FL ... Therefore, we need to find the optimum parameters of the ASM that best fit of the model to the target structure ...
Research Interests:
Research Interests:
In this paper we present methods for assessing the quality of facial images, degraded by blurring and facial expressions, for recognition. To assess the blurring effect, we measure the level of blurriness in the facial images by... more
In this paper we present methods for assessing the quality of facial images, degraded by blurring and facial expressions, for recognition. To assess the blurring effect, we measure the level of blurriness in the facial images by statistical analysis in the Fourier domain. Based on this analysis, a function is proposed to predict the performance of face recognition on blurred images. To assess facial images with expressions, we use Gaussian Mixture Models (GMMs) to represent images that can be recognized with the Eigenface method, we refer to these images as “Good Quality”, and images that cannot be recognized, we refer to these images as “Poor Quality”. During testing, we classify a given image into one of the two classes. We use the FERET and Cohn-Kanade facial image databases to evaluate our algorithms for image quality assessment. The experimental results demonstrate that the prediction function for assessing the quality of blurred facial images is successful. In addition, our experiments show that our approach for assessing facial images with expressions is successful in predicting whether an image has a good quality or poor quality for recognition. Although the experiments in this paper are based on the Eigenface technique, the assessment methods can be extended to other face recognition algorithms.
Research Interests: Image Processing, Pattern Recognition, Face Recognition, Statistical Analysis, Biometrics, and 15 moreImage Analysis, Facial expression, Image Quality, Quality Control, Fourier Analysis, Facial expressions, Image recognition, IS success, Density Estimation, Gaussian Mixture Model, Facies, Gaussian Mixture, Image Database, Assessment Methods, and Image Quality Assessment
ABSTRACT
Research Interests:
ABSTRACT This paper proposes and analyzes a novel FPGA-based System-on-Chip (SoC) module interconnection architecture called the Morphing Crossbar, which enables time-efficient partial dynamic reconfiguration of embedded systems built... more
ABSTRACT This paper proposes and analyzes a novel FPGA-based System-on-Chip (SoC) module interconnection architecture called the Morphing Crossbar, which enables time-efficient partial dynamic reconfiguration of embedded systems built with programmable logic. By combining local Crossbar and its companion peripheral interconnection bus, the Morphing Bus, compact embedded systems can be developed with both static and dynamic reconfigurability. The Morphing Crossbar decreases the overhead of interconnection and remapping by allowing the system components to be modularized into relocatable modules. This method increases the flexibility of dynamic partial reconfigurable system. The Morphing Crossbar, which allows adding, removing, or swapping module blocks inside the system on the fly, was implemented on the Xilinx Virtex-5 architecture. Our experimental results demonstrate that by using Morphing Crossbar, adding, removing, and swapping modules can be performed 3.6 times faster compared to architectures without Morphing Crossbar.
Research Interests:
ABSTRACT This paper presents an architecture for automating the reconfiguration of system deal with unforeseen situations, named ReFrESH, for distributed autonomous embedded systems which 1) supports both hardware and software... more
ABSTRACT This paper presents an architecture for automating the reconfiguration of system deal with unforeseen situations, named ReFrESH, for distributed autonomous embedded systems which 1) supports both hardware and software reconfiguration based on task-related functional requirements without disturbing system at runtime; 2) provides a type of Embedded Virtual Machine to facilitate components distribution across node boundaries; 3) generates optimal configurations dynamically based on non-functional requirements. The feasibility of ReFrESH and its management algorithms are evaluated for “visual servoing” of three miniature robot scenario. Moreover, one self-adaptive application is implemented to show the realistic performance of ReFrESH. The results demonstrate that ReFrESH can enable the system to handle various situations dynamically and decrease the complexity of self-adaptation.
Research Interests:
ABSTRACT Presents the results of a preliminary study on understanding how humanoid robots can successfully improve social and communication skills among children with Autism Spectrum Disorders (ASD). Children with ASD experience deficits... more
ABSTRACT Presents the results of a preliminary study on understanding how humanoid robots can successfully improve social and communication skills among children with Autism Spectrum Disorders (ASD). Children with ASD experience deficits in appropriate verbal and nonverbal communication skills including motor control, emotional facial expressions, eye-gaze attention, and joint attention. Studies have shown that positive feedback from the robot on the participants' performance is an effective way to encourage children with ASD to communicate more. Other studies have also examined the use of affect recognition based on psycho physiological responses to modify the behaviors during a robotic game. However, there is limited information on the utility of humanoid robots' positive.
Research Interests:
ABSTRACT A behvior recognition approach is proposed based on time-frequency analysis and machine learning techniques to identify Parkinson's disease (PD) patients' behviors using local field potential (LFP) signals... more
ABSTRACT A behvior recognition approach is proposed based on time-frequency analysis and machine learning techniques to identify Parkinson's disease (PD) patients' behviors using local field potential (LFP) signals obtained from a deep brain stimulation (DBS) system. Specifically, the amplitude-time-frequency-variance features are extracted by the matching pursuit decomposition (MPD) algorithm from LFP signals sampled by a DBS lead from the subthalamic (STN) area. Using the extracted feature vectors, different hidden Markov models (HMMs) including discrete and continuous HMMs are trained and then used to recognize different human behviors. The experiment results demonstrate the feasibility, effectiveness and accuracy of our proposed method.
Research Interests:
This paper presents a method to diagnose prostate cancer on multiparametric magnetic resonance imaging (Mp-MRI) using the shearlet transform. The objective is classification of benign and malignant regions on transverse relaxation time... more
This paper presents a method to diagnose prostate cancer on multiparametric magnetic resonance imaging (Mp-MRI) using the shearlet transform. The objective is classification of benign and malignant regions on transverse relaxation time weighted (T2W), dynamic contrast enhanced (DCE), and apparent diffusion coefficient (ADC) images. Compared with conventional wavelet filters, shearlet has inherent directional sensitivity, which makes it suitable for characterizing small contours of cancer cells. By applying a multi-scale decomposition, the shearlet transform captures visual information provided by edges detected at different orientations and multiple scales in each region of interest (ROI) of the images. ROIs are represented by histograms of shearlet coefficients (HSC) and then used as features in Support Vector Machines (SVM) to classify ROIs as benign or malignant. Experimental results show that our method can recognize carcinoma in T2W, DCE, and ADC with overall sensitivity of 92%...
Research Interests:
ABSTRACT
Research Interests:
This paper describes a novel approach to pattern classification that combines Parzen window and support vector machines. Pattern classification is usually performed in universes where all possible categories are defined. Most of the... more
This paper describes a novel approach to pattern classification that combines Parzen window and support vector machines. Pattern classification is usually performed in universes where all possible categories are defined. Most of the current supervised learning classification techniques do not account for undefined categories. In a universe that is only partially defined, there may be objects that do not fall
Research Interests:
ABSTRACT Some patients who have suffered a traumatic brain injury (TBI) or a spinal cord injury (SCI) lose the ability to operate a computer via traditional methods (mouse and keyboard) or common alternative methods (voice control). The... more
ABSTRACT Some patients who have suffered a traumatic brain injury (TBI) or a spinal cord injury (SCI) lose the ability to operate a computer via traditional methods (mouse and keyboard) or common alternative methods (voice control). The inability to move or speak makes it extremely difficult to communicate. Currently patients must rely on blinking yes or no to a series of questions that the caretaker asks in order to communicate their wants and needs. This system relies on the caretaker anticipating the patient’s needs in a timely manner which is not ideal. The purpose of this Senior Capstone Design project is to develop an eye tracking system to allow TBI and SCI patients to operate a computer exclusively using eye movements. Occupational Therapists at Craig Hospital, a recognized leader in TBI and SCI patient care, have provided the design team with access to a focus group of patients and insight into the needs and constraints of their patient population. With the ability to operate a computer, the patient will have the ability to more fully engage in communication by typing words on a screen for their caretakers to read and in a broader sense: e-mail and instant message family and friends as well as engage in social networking sites. The ability to control a computer allows for choices in entertainment from music to television to newspapers and magazines.
Research Interests:
Research Interests:
ABSTRACT In this paper, an image-based method is presented for fall detection using statistical human posture sequence modeling. Specifically, a series of laboratory simulated falls and activities of daily living (ADLs) are performed and... more
ABSTRACT In this paper, an image-based method is presented for fall detection using statistical human posture sequence modeling. Specifically, a series of laboratory simulated falls and activities of daily living (ADLs) are performed and recorded by a Kinect sensor as training video data. The skeleton view of a human body in these video recordings is extracted using the Kinect for Windows SDK. Hidden Markov Models are used for modeling the fall posture sequences and distinguishing different fall activities and ADLs. Our experimental results demonstrate an average fall recognition rate above 80% and the capability of early warning for falls.
Research Interests:
Research Interests:
Research Interests:
ABSTRACT Facial action unit (AU) detection is a challenging topic in computer vision and pattern recognition. Most existing approaches design classifiers to detect AUs individually or AU combinations without considering the intrinsic... more
ABSTRACT Facial action unit (AU) detection is a challenging topic in computer vision and pattern recognition. Most existing approaches design classifiers to detect AUs individually or AU combinations without considering the intrinsic relations among AUs. This paper presents a novel method, lp-norm multi-task multiple kernel learning (MTMKL), that jointly learns the classifiers for detecting the absence and presence of multiple AUs. lp-norm MTMKL is an extension of the regularized multi-task learning, which learns shared kernels from a given set of base kernels among all the tasks within Support Vector Machines (SVM). Our approach has several advantages over existing methods: (1) AU detection work is transformed to a MTL problem, where given a specific frame, multiple AUs are detected simultaneously by exploiting their inter-relations; (2) lp-norm multiple kernel learning is applied to increase the discriminant power of classifiers. Our experimental results on the CK+ and DISFA databases show that the proposed method outperforms the state-of-the-art methods for AU detection.