Computer Engineering Commons^™

Face Centered Image Analysis Using Saliency And Deep Learning Based Techniques, Rui Guo

Doctoral Dissertations

Image analysis starts with the purpose of configuring vision machines that can perceive like human to intelligently infer general principles and sense the surrounding situations from imagery. This dissertation studies the face centered image analysis as the core problem in high level computer vision research and addresses the problem by tackling three challenging subjects: Are there anything interesting in the image? If there is, what is/are that/they? If there is a person presenting, who is he/she? What kind of expression he/she is performing? Can we know his/her age? Answering these problems results in the saliency-based object detection, deep learning structured …

Intrinsically Motivated Exploration In Hierarchical Reinforcement Learning, Christopher M. Vigorito

Doctoral Dissertations

The acquisition of hierarchies of reusable skills is one of the distinguishing characteristics of human intelligence, and the learning of such hierarchies is an important open problem in computational reinforcement learning (RL). In humans, these skills are learned during a substantial developmental period in which individuals are intrinsically motivated to explore their environment and learn about the effects of their actions. The skills learned during this period of exploration are then reused to great effect later in life to solve many unfamiliar problems very quickly. This thesis presents novel methods for achieving such developmental acquisition of skill hierarchies in artificial …

Novel Approaches To Clustering, Biclustering Algorithms Based On Adaptive Resonance Theory And Intelligent Control, Sejun Kim

Doctoral Dissertations

"The problem of clustering is one of the most widely studied area in data mining and machine learning. Adaptive resonance theory (ART), an unsupervised learning clustering algorithm, is a clustering method that can learn arbitrary input patterns in a stable, fast and self-organizing way. This dissertation focuses on unsupervised learning methods, mostly based on variations of ART.

Hierarchical ART clustering is studied by generating a tree of ART units with GPU based parallelization to provide fast and finesse clustering. Experiment results show that the our method achieves significant training speed increase in generating deep ART trees compared with that from …

Clustering: Methodology, Hybrid Systems, Visualization, Validation And Implementation, Dao Minh Lam

Doctoral Dissertations

"Unsupervised learning is one of the most important steps of machine learning applications. Besides its ability to obtain the insight of the data distribution, unsupervised learning is used as a preprocessing step for other machine learning algorithm. This dissertation investigates the application of unsupervised learning into various types of data for many machine learning tasks such as clustering, regression and classification. The dissertation is organized into three papers. In the first paper, unsupervised learning is applied to mixed categorical and numerical feature data type to transform the data objects from the mixed type feature domain into a new sparser numerical …

Using Gpu To Accelerate Linear Computations In Power System Applications, Xue Li

Doctoral Dissertations

With the development of advanced power system controls, the industrial and research community is becoming more interested in simulating larger interconnected power grids. It is always critical to incorporate advanced computing technologies to accelerate these power system computations. Power flow, one of the most fundamental computations in power system analysis, converts the solution of non-linear systems to that of a set of linear systems via the Newton method or one of its variants. An efficient solution to these linear equations is the key to improving the performance of power flow computation, and hence to accelerating other power system applications based …

Hyperspectral Data Acquisition And Its Application For Face Recognition, Woon Cho

Doctoral Dissertations

Current face recognition systems are rife with serious challenges in uncontrolled conditions: e.g., unrestrained lighting, pose variations, accessories, etc. Hyperspectral imaging (HI) is typically employed to counter many of those challenges, by incorporating the spectral information within different bands. Although numerous methods based on hyperspectral imaging have been developed for face recognition with promising results, three fundamental challenges remain: 1) low signal to noise ratios and low intensity values in the bands of the hyperspectral image specifically near blue bands; 2) high dimensionality of hyperspectral data; and 3) inter-band misalignment (IBM) correlated with subject motion during data acquisition.

This dissertation …

Neuron Clustering For Mitigating Catastrophic Forgetting In Supervised And Reinforcement Learning, Benjamin Frederick Goodrich

Doctoral Dissertations

Neural networks have had many great successes in recent years, particularly with the advent of deep learning and many novel training techniques. One issue that has affected neural networks and prevented them from performing well in more realistic online environments is that of catastrophic forgetting. Catastrophic forgetting affects supervised learning systems when input samples are temporally correlated or are non-stationary. However, most real-world problems are non-stationary in nature, resulting in prolonged periods of time separating inputs drawn from different regions of the input space.

Reinforcement learning represents a worst-case scenario when it comes to precipitating catastrophic forgetting in neural networks. …

Arithmetic Logic Unit Architectures With Dynamically Defined Precision, Getao Liang

Doctoral Dissertations

Modern central processing units (CPUs) employ arithmetic logic units (ALUs) that support statically defined precisions, often adhering to industry standards. Although CPU manufacturers highly optimize their ALUs, industry standard precisions embody accuracy and performance compromises for general purpose deployment. Hence, optimizing ALU precision holds great potential for improving speed and energy efficiency. Previous research on multiple precision ALUs focused on predefined, static precisions. Little previous work addressed ALU architectures with customized, dynamically defined precision. This dissertation presents approaches for developing dynamic precision ALU architectures for both fixed-point and floating-point to enable better performance, energy efficiency, and numeric accuracy. These new …

Application Of Techniques For Map Estimation To Distributed Constraint Optimization Problem, Yoonheui Kim

Doctoral Dissertations

The problem of efficiently finding near-optimal decisions in multi-agent systems has become increasingly important because of the growing number of multi-agent applications with large numbers of agents operating in real-world environments. In these systems, agents are often subject to tight resource constraints and agents have only local views. When agents have non-global constraints, each of which is independent, the problem can be formalized as a distributed constraint optimization problem (DCOP). The DCOP is closely associated with the problem of inference on graphical models. Many approaches from inference literature have been adopted to solve DCOPs. We focus on the Max-Sum algorithm …

Privacy-Preserving Payments For Transportation Systems, Gesine Hinterwalder

Doctoral Dissertations

The operation of our society heavily relies on high mobility of people. Not only our social life but also our economy and trade are built upon a system where people need to be able to move around easily. The costs for building and maintaining a suitable transportation infrastructure to satisfy those needs are high, and to charge users is thus a central requirement. This calls for well functioning payment systems satisfying the multitude of requirements that transportation systems impose on them. Electronic payment systems have many benefits over traditional cash payments as they are easy to maintain, can be more …

Skybridge: A New Nanoscale 3-D Computing Framework For Future Integrated Circuits, Mostafizur Rahman

Doctoral Dissertations

Continuous scaling of CMOS has been the major catalyst in miniaturization of integrated circuits (ICs) and crucial for global socio-economic progress. However, continuing the traditional way of scaling to sub-20nm technologies is proving to be very difficult as MOSFETs are reaching their fundamental performance limits [1] and interconnection bottleneck is dominating IC operational power and performance [2]. Migrating to 3-D, as a way to advance scaling, has been elusive due to inherent customization and manufacturing requirements in CMOS architecture that are incompatible with 3-D organization. Partial attempts with die-die [3] and layer-layer [4] stacking have their own limitations [5]. We …

On Thermal Sensor Calibration And Software Techniques For Many-Core Thermal Management, Shiting Lu

Doctoral Dissertations

The high power density of a many-core processor results in increased temperature which negatively impacts system reliability and performance. Dynamic thermal management applies thermal-aware techniques at run time to avoid overheating using temperature information collected from on-chip thermal sensors. Temperature sensing and thermal control schemes are two critical technologies for successfully maintaining thermal safety. In this dissertation, on-line thermal sensor calibration schemes are developed to provide accurate temperature information. Software-based dynamic thermal management techniques are proposed using calibrated thermal sensors. Due to process variation and silicon aging, on-chip thermal sensors require periodic calibration before use in DTM. However, the calibration …

Energy Optimizations For Smart Buildings And Smart Grids, Aditya K. Mishra

Doctoral Dissertations

Modern buildings are heavy power consumers. For instance, of the total electricity consumed in the US, 75% is consumed in the residential and commercial buildings. This consumption is not evenly distributed over time. Typical consumption profile exhibits several peaks and troughs. The peakiness, in turn, dictates the electric grid's generation, transmission and distribution costs, and also the associated carbon emissions. This thesis discusses challenges involved in achieving the sustainability goals in buildings and electric grids. It investigates building and grid energy footprint optimization techniques to achieve the following goals: 1) making buildings energy efficient, 2) cutting building's electricity bills, 3) …

Design And Implementation Of An Economy Plane For The Internet, Xinming Chen

Doctoral Dissertations

The Internet has been very successful in supporting many network applications. As the diversity of uses for the Internet has increased, many protocols and services have been developed by the industry and the research community. However, many of them failed to get deployed in the Internet. One challenge of deploying these novel ideas in operational network is that the network providers need to be involved in the process. Many novel network protocols and services, like multicast and end-to-end QoS, need the support from network providers. However, since network providers are typically driven by business reasons, if they can not get …

Energy-Efficient Content Delivery Networks, Vimal Mathew

Doctoral Dissertations

Internet-scale distributed systems such as content delivery networks (CDNs) operate hundreds of thousands of servers deployed in thousands of data center locations around the globe. Since the energy costs of operating such a large IT infrastructure are a significant fraction of the total operating costs, we argue for redesigning them to incorporate energy optimization as a first-order principle. We focus on CDNs and demonstrate techniques to save energy while meeting client-perceived service level agreements (SLAs) and minimizing impact on hardware reliability. Servers deployed at individual data centers can be switched off at low load to save energy. We show that …

On Applications Of Relational Data, Samamon Khemmarat

Doctoral Dissertations

With the advances of technology and the popularity of the Internet, a large amount of data is being generated and collected. Much of these data is relational data, which describe how people and things, or entities, are related to one another. For example, data from sale transactions on e-commerce websites tell us which customers buy or view which products. Analyzing the known relationships from relational data can help us to discover knowledge that can benefit businesses, organizations, and our lives. For instance, learning the products that are commonly bought together allows businesses to recommend products to customers and increase their …

Physically Equivalent Intelligent Systems For Reasoning Under Uncertainty At Nanoscale, Santosh Khasanvis

Doctoral Dissertations

Machines today lack the inherent ability to reason and make decisions, or operate in the presence of uncertainty. Machine-learning methods such as Bayesian Networks (BNs) are widely acknowledged for their ability to uncover relationships and generate causal models for complex interactions. However, their massive computational requirement, when implemented on conventional computers, hinders their usefulness in many critical problem areas e.g., genetic basis of diseases, macro finance, text classification, environment monitoring, etc. We propose a new non-von Neumann technology framework purposefully architected across all layers for solving these problems efficiently through physical equivalence, enabled by emerging nanotechnology. The architecture builds …

Threat Analysis, Countermeaures And Design Strategies For Secure Computation In Nanometer Cmos Regime, Raghavan Kumar

Doctoral Dissertations

Advancements in CMOS technologies have led to an era of Internet Of Things (IOT), where the devices have the ability to communicate with each other apart from their computational power. As more and more sensitive data is processed by embedded devices, the trend towards lightweight and efficient cryptographic primitives has gained significant momentum. Achieving a perfect security in silicon is extremely difficult, as the traditional cryptographic implementations are vulnerable to various active and passive attacks. There is also a threat in the form of "hardware Trojans" inserted into the supply chain by the untrusted third-party manufacturers for economic incentives. Apart …

Data Security And Privacy In Smart Grid, Yue Tong

Doctoral Dissertations

This dissertation explores novel data security and privacy problems in the emerging smart grid.

The need for data security and privacy spans the whole life cycle of the data in the smart grid, across the phases of data acquisition, local processing and archiving, collaborative processing, and finally sharing and archiving. The first two phases happen in the private domains of an individual utility company, where data are collected from the power system and processed at the local facilities. When data are being acquired and processed in the private domain, data security is the most critical concern. The key question is …

A Magnetic Actuated Fully Insertable Robotic Camera System For Single Incision Laparoscopic Surgery, Xiaolong Liu

Doctoral Dissertations

Minimally Invasive Surgery (MIS) is a common surgical procedure which makes tiny incisions in the patients anatomy, inserting surgical instruments and using laparoscopic cameras to guide the procedure. Compared with traditional open surgery, MIS allows surgeons to perform complex surgeries with reduced trauma to the muscles and soft tissues, less intraoperative hemorrhaging and postoperative pain, and faster recovery time. Surgeons rely heavily on laparoscopic cameras for hand-eye coordination and control during a procedure. However, the use of a standard laparoscopic camera, achieved by pushing long sticks into a dedicated small opening, involves multiple incisions for the surgical instruments. Recently, single …

Computational Analysis Of Neutron Scattering Data, Benjamin Walter Martin

Doctoral Dissertations

This work explores potential methods for use in the detection and classification of defects within crystal structures via analysis of diffuse scattering data generated by single crystal neutron scattering experiments. The proposed defect detection methodology uses machine learning and image processing techniques to perform image texture analysis on neutron diffraction patterns generated by neutron scattering simulations. Once the methodology is presented, it is tested via a series of defect detection problems of increasing difficulty which utilize neutron scattering data simulated by a number of simulation techniques. As the problem difficulty is increased, the defect detection methodology is refined in order …

Compressed Sensing In Resource-Constrained Environments: From Sensing Mechanism Design To Recovery Algorithms, Shuangjiang Li

Doctoral Dissertations

Compressed Sensing (CS) is an emerging field based on the revelation that a small collection of linear projections of a sparse signal contains enough information for reconstruction. It is promising that CS can be utilized in environments where the signal acquisition process is extremely difficult or costly, e.g., a resource-constrained environment like the smartphone platform, or a band-limited environment like visual sensor network (VSNs). There are several challenges to perform sensing due to the characteristic of these platforms, including, for example, needing active user involvement, computational and storage limitations and lower transmission capabilities. This dissertation focuses on the study of …

Dynamic Simulation And Neuromuscular Control Of Movement: Applications For Predictive Simulations Of Balance Recovery, Misagh Mansouri Boroujeni

Doctoral Dissertations

Balance is among the most challenging tasks for patients with movement disorders. Study and treatment of these disorders could greatly benefit from combined software tools that offer better insights into neuromuscular biomechanics, and predictive capabilities for optimal surgical and rehabilitation treatment planning. A platform was created to combine musculoskeletal modeling, closed-loop forward dynamic simulation, optimization techniques, and neuromuscular control system design. Spinal (stretch-reflex) and supraspinal (operational space task-based) controllers were developed to test simulation-based hypotheses related to balance recovery and movement control. A corrective procedure (rectus femoris transfer surgery) was targeted for children experiencing stiff-knee gait and how this procedure …

Computational Framework For Small Animal Spect Imaging: Simulation And Reconstruction, Sang Hyeb Lee

Doctoral Dissertations

Small animal Single Photon Emission Computed Tomography (SPECT) has been an invaluable asset in biomedical science since this non-invasive imaging technique allows the longitudinal studies of animal models of human diseases. However, the image degradation caused by non-stationary collimator-detector response and single photon emitting nature of SPECT makes it difficult to provide a quantitative measure of 3D radio-pharmaceutical distribution inside the patient. Moreover, this problem exacerbates when an intra-peritoneal X-ray contrast agent is injected into a mouse for low-energy radiotracers.

In this dissertation, we design and develop a complete computational framework for the entire SPECT scan procedure from the radio-pharmaceutical …

Universal Schema For Knowledge Representation From Text And Structured Data, Limin Yao

Doctoral Dissertations

In data integration we transform information from a source into a target schema. A general problem in this task is loss of fidelity and coverage: the source expresses more knowledge than that can be fit into the target schema, or knowledge that is hard to fit into any schema at all. This problem is taken to an extreme in information extraction (IE) where the source is natural language---one of the most expressive forms of knowledge representation. To address this issue, one can either automatically learn a latent schema emergent in text (a brittle and ill-defined task), or manually define schemas. …

An Opportunistic Service Oriented Approach For Robot Search, Dan Xie

Doctoral Dissertations

Health care for the elderly poses a major challenge as the baby boomer generation ages. Part of the solution is to develop technology using sensor networks and service robotics to increase the length of time that an elder can remain at home. Since moderate immobility and memory impairment are common as people age, a major problem for the elderly is locating and retrieving frequently used "common" objects such as keys, cellphones, books, etc. However, for robots to assist people while they search for objects, they must possess the ability to interact with the human client, complex client-side environments and heterogeneous …

Managing And Leveraging Variations And Noise In Nanometer Cmos, Vikram B. Suresh

Doctoral Dissertations

Advanced CMOS technologies have enabled high density designs at the cost of complex fabrication process. Variation in oxide thickness and Random Dopant Fluctuation (RDF) lead to variation in transistor threshold voltage V_th. Current photo-lithography process used for printing decreasing critical dimensions result in variation in transistor channel length and width. A related challenge in nanometer CMOS is that of on-chip random noise. With decreasing threshold voltage and operating voltage; and increasing operating temperature, CMOS devices are more sensitive to random on-chip noise in advanced technologies. In this thesis, we explore novel circuit techniques to manage the impact of …

Learning Parameterized Skills, Bruno Castro Da Silva

Doctoral Dissertations

One of the defining characteristics of human intelligence is the ability to acquire and refine skills. Skills are behaviors for solving problems that an agent encounters often—sometimes in different contexts and situations—throughout its lifetime. Identifying important problems that recur and retaining their solutions as skills allows agents to more rapidly solve novel problems by adjusting and combining their existing skills. In this thesis we introduce a general framework for learning reusable parameterized skills. Reusable skills are parameterized procedures that—given a description of a problem to be solved—produce appropriate behaviors or policies. They can be sequentially and hierarchically combined with other …

On The Deployment Of On-Chip Noise Sensors, Tao Wang

Doctoral Dissertations

"The relentless technology scaling has led to significantly reduced noise margin and complicated functionalities. As such, design time techniques per se are less likely to ensure power integrity, resulting in runtime voltage emergencies. To alleviate the issue, recently several works have shed light on the possibilities of dynamic noise management systems. Most of these works rely on on-chip noise sensors to accurately capture voltage emergencies. However, they all assume that the placement of the sensors is given. It remains an open problem in the literature how to optimally place a given number of noise sensors for best voltage emergency detection. …

Cormem Digital Reasoning Architecture Using Cmos Technology, Indira Priyadarshini Dugganapally

Doctoral Dissertations

”This dissertation describes the application of a multi-level, memory-based approach for building digital circuits. To reflect the alternative approach, the basic science is termed digital reasoning and the specific CorMem technology is based on recent patents. CMOS transistors are used in a non-traditional way for multi-level operations and memory manipulation. The combination of multi-level architectures and matrix algebra principles can create flexible, modular systems using standard fabrication methods and can avoid many of the limitations of other multi-valued logic approaches.

Quaternary, memory-based systems are developed to implement logic-gate-type functions, digital adder circuits, a complete arithmetic and logic unit (ALU), quaternary-to-binary …