Recently, several approaches for pedestrian detection have been investigated using discriminatively trained part based models with which Histogram of Oriented Gradients (HOG) showed to be a robust feature. In this paper, we propose a new feature based on HOG to be used with the discriminatively trained part framework for pedestrian detection. Our method is based on computing the image mixed partial derivatives to be used to redefine the gradients of some pixels and to reweigh the vote at all pixels with respect to the original HOG. Our approach was tested on the PASCAL2007 and INRIA person dataset and showed to have an outstanding performance.

Deformable models are common in image modeling and analysis. Random objects provide major challenges as shapes and appearances are hard to quantify; hence, formulation of deformable models are much harder to construct and validate. In this paper, we examine the effect of randomness on building the shape and appearance models for small-size lung nodules (≤ 1cm) which appear in computed tomography (CT) of the human chest. We devise an approach for annotation, which lends a standard mechanism for building traditional active appearance (AAM), active shape (ASM) and active tensor models (ATM). We illustrate the effectiveness of AAM for nodule detection.

In this paper, We propose a new approach for solving the Lambertian shape from shading (SFS) problem. The proposed technique employs image information as well as camera calibration parameters. We claim that the incorporation of camera parameters results in a better surface reconstruction. In this work, we slightly modify this more realistic model in order to take into account the attenuation of the illumination due to distance. We propose a new variational formulation that relates an evolving surface model with image information, taking into consideration that the image is taken by a perspective camera with known parameters. A new energy functional is formulated to incorporate brightness, smoothness, and integrability constraints. All of these terms assume a hyper surface that evolves in time to meet their criteria. Gradient descent optimization with Euler-Lagrange is used for optimization. We constructed a database of real surfaces as a ground-truth for the experiments. The proposed approach shows superior performance when compared with other approaches in the literature.

Accurate 3D modeling of the human teeth/jaw helps patients avoid the discomfort of the mold process, and improves the data accuracy for oral orthodontists and dental care personnel. Since the surface of the human tooth is almost textureless, Shape from Shading (SFS) has been successfully adopted in solving this problem. In this paper, we attempt to improve the limitations of previous 3D tooth reconstruction algorithms by developing a new approach for shape reconstruction from single image shading with Two-Dimensional Principle Component Analysis (2D-PCA) shape priors. The surface reflectance is modeled using Oren-Nayar-Wolff model which accounts for the retro-reflection case and we experimentally prove that the teeth surface follows the microfacet theory. Our formulation exploits the shape priors as extracted from a set of training CT scans of real human jaws. Our experiments provide promising quantitative metric results for the proposed approach. This work is fundamental for establishing an optical-based approach for reconstructing the human jaw that is inexpensive and does not use ionizing radiation.

In this paper we propose an educational robotic system for teaching kids at nursery schools how to write and draw simple shapes. Our system uses the humanoid robot NAO to draw shapes appearing on a computer screen. The system uses light polarization property for fast detection of the screen despite of its content. We also propose a mapping from the image domain to the robot space.

A new variational level set approach is proposed for lung nodule segmentation in lung CT scans. A general lung nodule shape model is proposed using implicit spaces as a signed distance function. The shape model is fused with the image intensity statistical information in a variational segmentation framework. The nodule shape model is mapped to the image domain by a global transformation that includes inhomogeneous scales, rotation, and translation parameters. A matching criteria between the shape model and the image implicit representations is employed to handle the alignment process. Transformation parameters evolve through gradient descent optimization to handle the shape alignment process and hence mark the boundaries of the nodule “head.” The embedding process takes into consideration the image intensity as well as prior shape information. A nonparametric density estimation approach is employed to handle the statistical intensity representation of the nodule and background regions. The proposed technique does not depend on nodule type or location. Exhaustive experimental and validation results are demonstrated on 742 nodules obtained from four different CT lung databases, illustrating the robustness of the approach.

Lung nodules from low dose CT (LDCT) scans may be used for early detection of lung cancer. However, these nodules vary in size, shape, texture, location, and may suffer from occlusion within the tissue. This paper presents an approach for segmentation of lung nodules detected by a prior step. First, regions around the detected nodules are segmented; using automatic seed point placement levels sets. The outline of the nodule region is further improved using the curvature characteristics of the segmentation boundary. We illustrate the effectiveness of this method for automatic segmentation of the Juxta-pleural nodules.

Supervisory control and data acquisition (SCADA) systems are critical to today's industrial facilities and infrastructures. SCADA systems have evolved into large and complex networks of information systems and are increasingly vulnerable to various types of cyber-security risks. Identifying and managing risks in SCADA systems has become critical in ensuring the safety and reliability of these facilities and infrastructures. Most of the existing research on SCADA risk modeling and management has focused on probability-based or quantitative approaches. While probabilistic approaches have proven to be useful, they also suffer from common problems such as simplifying assumptions, large implementation costs, and inability to completely capture all the important aspects of risk. This paper proposes a digraph model for SCADA systems that allows formal, explicit representation of a SCADA system. A number of risk management methods are presented and discussed for a SCADA system based on the proposed model. The methods are applied to a chemical distillation application as a case study, and shows promising initial results in identifying areas of system vulnerability.

This work presents a novel approach in automatic detection of the lung nodules and is compared with respect to parametric nodule models in terms of sensitivity and specificity. A Statistical method is used for generating data driven models of the nodules appearing in low dose CT (LDCT) scans of the human chest. Four types of common lung nodules are analyzed using the Procrustes based AAM method to create descriptive lung nodules. Performance of the new nodule models on clinical datasets is significant over parametric nodule models in both sensitivity and specificity. The new nodule modeling approach is also applicable for automatic classification of nodules into pathologies given a descriptive database. This approach is a major step forward for early diagnosis of lung cancer.

This paper deals with segmentation of the lung tissues from low dose CT (LDCT) scans of the chest. Goal is correct segmentation as well as maintaining the details of the lung region in the chest cavity. In particular, it is essential that the lung nodules inside the lungs as well as on the boundary regions be maintained for subsequent steps that aim at automatic detection and classification of nodules from LDCT scanning; a step for early diagnosis of lung cancer. An approach for segmentation based on combination of EM algorithm and morphological operations is presented. This algorithm is compared with respect two other approaches that are based on level sets and energy optimization by the Graph Cuts technique. Performance evaluation is conducted on a labeled data set from the Early Lung Cancer Action Program (ELCAP) database. The new segmentation approach provides comparable results to Level sets and Graph-Cuts, with the advantage of faster execution time, and minimal user interception.

The quality of the lung nodule models determines the success of lung nodule detection. This paper describes aspects of our data-driven approach for modeling lung nodules using the texture and shape properties of real nodules to form an average model template per nodule type. The ELCAP low dose CT (LDCT) scans database is used to create the required statistics for the models based on modern computer vision techniques. These models suit various machine learning approaches for nodule detection including Bayesian methods, SVM and Neural Networks, and computations may be enhanced through genetic algorithms and Adaboost. The eminence of the new nodule models are studied with respect to parametric models showing significant improvements in both sensitivity and specificity.

One critical threat facing many organizations is the inside attacks from masqueraders, internal users or external intruders who exploit legitimate user identity and perform malicious attacks. Anomaly intrusion detection systems can be deployed to build a user behavior profile from his/her past activities in a computer system and detect masqueraders if a large deviation is observed. In this paper, we use a finite automata based model to construct a normal behavior reference model from the analysis of shell command sequences. A fuzzy evaluation mechanism is proposed to classify the degree of threat as linguistic terms. The fuzzy number calculated from the output of a fuzzy inference system is compared with predefined generalized fuzzy numbers representing different threat levels. A case will be labeled as the linguistic term which has the highest similarity value with it. Experiments conducted on two data sets both achieved high detection rates of masqueraders and few false alarms, which stand out other methods.

This paper presents a brief overview of current work at the University of Louisville toward the use of systems engineering methodologies for control of the software engineering process. While this work is still in its initial stages, the methodologies for failure modes, effects and criticality analysis seem to be particularly appropriate. Two medium size software applications of these methodologies are currently under investigation in our laboratory.

Despite extensive efforts during recent years within the technical community to improve computer security, serious security problems continue to receive increasing coverage in both the popular and technical media. A large part of the problem stems from poor systems engineering of software and networks. In biological systems, natural internal immune system responses identify and protect the organism. A key mechanism in this immunity process is the ability to distinguish between self (i. e. normal organisms or behaviors) and non-self (i. e. abnormal or anomalous behavior). To deal with the ambiguities and imprecision in the process of anomaly detection for computer system security, we introduce a hierarchical fuzzy inference system to capture normal behavior deviations. Fuzzy logic has been widely used in control systems, decision-making, information retrieval, and many other applications. In this paper, we explore its capability in the area of computer security threat evaluation modeling and anomaly detection. Initial studies indicate promising results for this approach.

The rapid diagnosis of faults in computerized manufacturing systems is crucial to reduce expensive downtime. Many hard computing approaches use either symptom-based or functional reasoning. Symptom-based approaches are unable to handle exceptions, while functional approaches are computationally expensive, and, thus unable to produce a real-time response. Current hybrid approaches which combines the two hard computing methods are too structured in their approach to switching between reasoning methods and, thus fail to provide rapid response comparable to humans. This paper presents a robust, extensible approach to fault diagnosis combines these hard computing methods with the soft computing of agents using fuzzy logic. This fusion of hard and soft computing methods allows unstructured switching between reasoning methods by utilizing multiple intelligent agents which examine the problem domain from a variety of perspectives.

Robot safety is a critical and largely unsolved problem involving the interaction of man and machine. The paper presents a new approach to robot safety which uses an integrated sensing architecture for monitoring the robot workspace, and a new detection and decision logic for regulating the safe operation of the robot. Sensory information is fused through a trained neural network to produce a map of the hazards. Using this combined map, and information about the robot's current position and velocity, a set of fuzzy logic rules has been implemented to regulate robot activity. Simulation results presented in the paper indicate that this method is both effective in detection of potentially hazardous situations and computationally feasible.

When faults in a device occur, the subdevices indicated to be at fault may not be the sources of failure. The ability of a diagnostic system to trace and locate the true sources of failure is therefore very important. This paper presents new algorithms for locating failure sources by using knowledge of device structure and fault propagation paths, and through the use of sequential testing of system subdevices. The inference strategy presented in this paper uses a systematic method of generating subdevices for testing as a way of reducing the candidate space for both single fault and multiple fault situations. In the case of a single failure source, the approach presented will correctly identify the subdevice which is the failure source. In the case of multiple failure sources, a small set of possible failure sources will be identified whose collective failure will explain all the known abnormalities.>

A description is given of the object-oriented analysis and implementation of a knowledge-based system for diagnosis of computer integrated manufacturing cells using a hybrid reasoning approach. The functional hierarchy of the manufacturing cell forms an inheritance hierarchy in multistrata format for the deep reasoning component, which leads to a set of symptom oriented production rules for final fault isolation and identification. Search path selection and computation are performed by a set of methods associated with the inheritance hierarchies.>

A computerized system is presented to provide expert assistance to a physician in the evaluation of electromyographic findings in the clinical diagnosis of compressive radiculopathies. The system uses an object-oriented, frame-based representation of the nerve-root and muscular structure of the lower back and leg, and uses a rule-based reasoning system to interpret electromyographic findings and relate them to potential pathologies in the nerve roots in the lower spinal column. A novel feature of this research was the development of a hybrid system of Bayesian regulated belief functions and symbolic endorsements for the resolution of uncertainty in the clinical observations. This hybrid system provided quantitative estimates of the nerve pathologies. while simultaneously providing the physician with qualitative information regarding the interrelations of the clinical findings and the diagnosis.>

The author provides an overview of recent research into specially designed computer architectures which can provide enhanced computational capabilities to meet the special needs of real-time control of robots and other automation systems. He reviews standard techniques for improving computational power through parallelism and special hardware and then considers how these approaches can be applied to problems in robot control, sensing, and coordination. It is noted that parallel and concurrent computing architectures offer several promising routes for achieving performance improvements, but they must meet the criteria of practical implementability, including connectability, performance, and cost. Some proposed architectures for performance improvements are surveyed, several of which have already achieved some degree of implementation. This survey is intended both to highlight significant progress to date and to provide an overview of the emerging trends for the development of specialized computer architectures in this field of application.>

Major areas of research currently under investigation in the development of safe and reliable robots are summarized. These areas include human factors issues, the design of the robot itself, systems engineering issues and standards development. It is concluded that much more research needs to be done in preparation for more sophisticated robot applications.>

Optimal state estimation procedures, such as the Kalman-Bucy filter, require a high-speed parallel implementation to fully realize their potential. This correspondence first analyzes and restructures these equations based on an expression tree representation. The resulting equations are then simulated on three proposed SIMD architectures (rectangular systolic array, linear array, and quad-clustered tree array) and evaluated for speedup, efficiency, and utilization as compared to a known serial implementation. Results are presented which graphically show the design tradeoffs between these architectures.

The advantages of using probabilistic models for the uncertainties encountered in manipulator path planning are presented.

The development of a new hierarchical linguistic-based learning control structure for complex systems is described. The complete system will be able to interact with a human operator in a high-level language at the highest level of the hierarchy and will be able to control detailed motions of a complex physical system at the lowest level of the hierarchy. Each level of the hierarchy is defined by a formal grammar which can generate exactly the class of admissible control actions at that level. A new linguistic structure, the linguistic decision schema, is proposed to specify the mapping between linguistic elements in adjacent levels. In the most general form, the decision schema incorporates a learning algorithm to obtain asymptotically optimal mappings for control under stochastic environments.