Engineering Commons^™

Reasoning Across Language And Vision In Machines And Humans, Andrei Barbu

Open Access Dissertations

Humans not only outperform AI and computer-vision systems, but use an unknown computational mechanism to perform tasks for which no suitable approaches exist. I present work investigating both novel tasks and how humans approach them in the context of computer vision and linguistics. I demonstrate a system which, like children, acquires high-level linguistic knowledge about the world. Robots learn to play physically-instantiated board games and use that knowledge to engage in physical play. To further integrate language and vision I develop an approach which produces rich sentential descriptions of events depicted in videos. I then show how to simultaneously detect …

Improving Multicore Resource Efficiency And Performance, Syed Ali Raza Jafri

Open Access Dissertations

With clock speeds stagnating for the last few years and multi-cores having replaced uniprocessors, software development must now turn towards shared memory parallel programming to continue enhancing performance. Shared memory parallel programming; however is significantly more challenging than its sequential counterpart. Conventional shared memory parallel programs can fall victim to deadlocks, livelocks and data races which are hard to detect and debug. Aside from programming complexity chipmultiprocessors need a scalable, low latency, high bandwidth interconnect fabric to deliver performance. Conventional interconnects such as crossbars and buses can deliver low latency but do not scale with increasing number of cores. Researchers …

Automatically Optimizing Tree Traversal Algorithms, Youngjoon Jo

Open Access Dissertations

Many domains in computer science, from data-mining to graphics to computational astrophysics, focus heavily on irregular applications. In contrast to regular applications, which operate over dense matrices and arrays, irregular programs manipulate and traverse complex data structures like trees and graphs. As irregular applications operate on ever larger datasets, their performance suffers from poor locality and parallelism. Programmers are burdened with the arduous task of manually tuning such applications for better performance. Generally applicable techniques to optimize irregular applications are highly desired, yet scarce.

In this dissertation, we argue that, for an important subset of irregular programs which arises in …

Resource-Aware Distributed Particle Filtering For Cluster-Based Object Tracking In Wireless Camera Networks, Kihyun Hong

Open Access Dissertations

The proliferation of miniaturized low-power computing devices, advances in wireless communications, and the availability of inexpensive imaging sensors have enabled the development of wireless camera networks (WCN). In this dissertation, we consider the problem of real-time object tracking with a WCN. Existing object tracking methods designed for multi-camera systems do not take into account the unique constraints of WCNs. Specifically, an effective object tracking system for WCNs must anticipate unreliable network communication, limited memory, and limited computational power in each camera node. In particular, unreliable communication degrades the quality of the visual information shared by the cameras, which ultimately degrades …

Source Code Retrieval From Large Software Libraries For Automatic Bug Localization, Bunyamin Sisman

Open Access Dissertations

This dissertation advances the state-of-the-art in information retrieval (IR) based approaches to automatic bug localization in software. In an IR-based approach, one first creates a search engine using a probabilistic or a deterministic model for the files in a software library. Subsequently, a bug report is treated as a query to the search engine for retrieving the files relevant to the bug. With regard to the new work presented, we first demonstrate the importance of taking version histories of the files into account for achieving significant improvements in the precision with which the files related to a bug are located. …

Transition Faults And Transition Path Delay Faults: Test Generation, Path Selection, And Built-In Generation Of Functional Broadside Tests, Bo Yao

Open Access Dissertations

As the clock frequency and complexity of digital integrated circuits increase rapidly, delay testing is indispensable to guarantee the correct timing behavior of the circuits. In this dissertation, we describe methods developed for three aspects of delay testing in scan-based circuits: test generation, path selection and built-in test generation.

We first describe a deterministic broadside test generation procedure for a path delay fault model named the transition path delay fault model, which captures both large and small delay defects. Under this fault model, a path delay fault is detected only if all the individual transition faults along the path are …

Predictive Duty Cycling Of Radios And Cameras Using Augmented Sensing In Wireless Camera Networks, Joonhwa Shin

Open Access Dissertations

Energy efficiency dominates practically every aspect of the design of wireless camera networks (WCNs), and duty cycling of radios and cameras is an important tool for achieving high energy efficiencies. However, duty cycling in WCNs is made complex by the camera nodes having to anticipate the arrival of the objects in their field-of-view. What adds to this complexity is the fact that radio duty cycling and camera duty cycling are tightly coupled notions in WCNs.

Abstract In this dissertation, we present a predictive framework to provide camera nodes with an ability to anticipate the arrival of an object in the …

Ant: A Framework For Increasing The Efficiency Of Sequential Debugging Techniques With Parallel Programs, Jae-Woo Lee

Open Access Dissertations

Bugs in sequential programs cost the software industry billions of dollars in lost productivity each year. Even if simple parallel programming models are created, they will not reduce the level of sequential bugs in programs below that of sequential programs. It can be argued that the complexity of current parallel programming models may increase the number of sequential bugs in parallel programs because they distract the programmer from the core logic of the program.

Tools exist that identify statements related to sequential bugs and allow those bugs to be more quickly located and fixed. Their use in parallel programs will …

Mitigating The Cost, Performance, And Power Overheads Induced By Load Variations In Multicore Cloud Servers, Yu-Ju Hong

Open Access Dissertations

Load variations whether in space or time pose a significant challenge to system designers. These load variations may induce inefficiencies such as load imbalance and over-provisioning, resulting in performance/power/cost overheads. The goal of my research is to mitigate such variation-induced overheads in multicore cloud servers.

First, I focus on power/performance overheads in on-chip networks of a multicore chip. We design an on-chip network that is robust in both performance and energy across applications for time- and space-varying loads. Existing flow control mechanisms that perform well at high (low) loads suffer power and/or energy overheads at low (high) loads. In contrast, …

Hypothesize-And-Verify Based Solutions For Place Recognition And Mobile Robot Self-Localization In Interior Hallways, Khalil Mustafa Ahmad Yousef

Open Access Dissertations

There is much research interest currently in having mobile robots build accurate and visually dense models ofinterior space as they traverse through such spaces. One of the interesting problems that has came out of this research is that of visual place recognition and self-localization. This is the problem that forms the focus of the present dissertation. We show how dense and accurate 3D models of the interior space can be constructed using a hierarchical sensor-fusion architecture. Our system fuses images from a single photometric camera with range data from a laser scanning sensor. The range data used is rudimentary--the range …

Semantically Grounded Learning From Unstructured Demonstrations, Scott D. Niekum

Open Access Dissertations

Robots exhibit flexible behavior largely in proportion to their degree of semantic knowledge about the world. Such knowledge is often meticulously hand-coded for a narrow class of tasks, limiting the scope of possible robot competencies. Thus, the primary limiting factor of robot capabilities is often not the physical attributes of the robot, but the limited time and skill of expert programmers. One way to deal with the vast number of situations and environments that robots face outside the laboratory is to provide users with simple methods for programming robots that do not require the skill of an expert.

For this …

Reconfigurable Technologies For Next Generation Internet And Cluster Computing, Deepak C. Unnikrishnan

Open Access Dissertations

Modern web applications are marked by distinct networking and computing characteristics. As applications evolve, they continue to operate over a large monolithic framework of networking and computing equipment built from general-purpose microprocessors and Application Specific Integrated Circuits (ASICs) that offers few architectural choices. This dissertation presents techniques to diversify the next-generation Internet infrastructure by integrating Field-programmable Gate Arrays (FPGAs), a class of reconfigurable integrated circuits, with general-purpose microprocessor-based techniques. Specifically, our solutions are demonstrated in the context of two applications - network virtualization and distributed cluster computing.

Network virtualization enables the physical network infrastructure to be shared among several …

Hybrid Radio Frequency And Video Framework For Identity-Aware Augmented Perception In Disaster Management And Assistive Technologies, Xunyi Yu

Open Access Dissertations

In this dissertation, we introduce a hybrid radio frequency and video framework that enables identity aware-augmented perception. Identity-aware augmented perception enhances users' perception of the surrounding by collecting and analyzing information pertaining to each identifiable or tractable target nearby aggregated from various sensors, and presents it visually or audibly augmenting users' own sensory perceptions. We target two application areas of disaster management and assistive technologies. Incident commanders and first responders can use the technology to perceive information specific to a victim, e.g. triage level, critical conditions, visually superimposed on third person or first person video. The blind and visually impaired …

Online Management Of Resilient And Power Efficient Multicore Processors, Rance Rodrigues

Open Access Dissertations

The semiconductor industry has been driven by Moore's law for almost half a century. Miniaturization of device size has allowed more transistors to be packed into a smaller area while the improved transistor performance has resulted in a significant increase in frequency. Increased density of devices and rising frequency led, unfortunately, to a power density problem which became an obstacle to further integration. The processor industry responded to this problem by lowering processor frequency and integrating multiple processor cores on a die, choosing to focus on Thread Level Parallelism (TLP) for performance instead of traditional Instruction Level Parallelism (ILP).

While …