Open Access. Powered by Scholars. Published by Universities.®
- Discipline
-
- Physical Sciences and Mathematics (825)
- Computer Engineering (733)
- Computer Sciences (730)
- Mechanical Engineering (376)
- Biomedical Engineering and Bioengineering (211)
-
- Electrical and Computer Engineering (127)
- Chemical Engineering (88)
- Life Sciences (76)
- Materials Science and Engineering (62)
- Civil and Environmental Engineering (53)
- Bioimaging and Biomedical Optics (51)
- Environmental Engineering (51)
- Electrical and Electronics (49)
- Physics (47)
- Aerospace Engineering (42)
- Medicine and Health Sciences (35)
- Engineering Science and Materials (34)
- Optics (34)
- Mechanics of Materials (30)
- Chemistry (26)
- Nanoscience and Nanotechnology (25)
- Aerodynamics and Fluid Mechanics (24)
- Neuroscience and Neurobiology (24)
- Molecular, Cellular, and Tissue Engineering (17)
- Biochemistry, Biophysics, and Structural Biology (16)
- Biomedical (16)
- Artificial Intelligence and Robotics (15)
- Medical Specialties (15)
- Applied Mathematics (14)
- Keyword
-
- CFD (17)
- Machine learning (17)
- Optimization (16)
- Machine Learning (11)
- Microfluidics (10)
-
- Fluorescence (9)
- Imaging (9)
- Mems411 (9)
- Biomaterials (8)
- Cancer (8)
- FPGA (8)
- Image reconstruction (8)
- Polarization (8)
- Aerosol (7)
- Biomechanics (7)
- Cyanobacteria (6)
- Deep Learning (6)
- Deep learning (6)
- Lignin (6)
- Mechanobiology (6)
- Nanoparticles (6)
- Nucleation (6)
- Acoustofluidics (5)
- Computational Fluid Dynamics (5)
- Computational imaging (5)
- Computed tomography (5)
- Control (5)
- Focused ultrasound (5)
- Metabolic Engineering (5)
- Metabolic engineering (5)
- Publication Year
- Publication
-
- McKelvey School of Engineering Graduate Student Theses & Dissertations (783)
- All Computer Science and Engineering Research (683)
- Mechanical Engineering Design Project Class (242)
- Washington University / UMSL Mechanical Engineering Design Project JME 4110 (40)
- All Theses and Dissertations (ETDs) (36)
-
- Electrical & Systems Engineering Publications and Presentations (23)
- Computer Science and Engineering Faculty Research (3)
- Electrical and Systems Engineering Capstone Design Projects (3)
- Generative AI Teaching Activities (3)
- Undergraduate Research Symposium Posters (3)
- Olin Business School Graduate Student Theses and Dissertations (2)
- Books and Monographs (1)
- Doctor of Business Administration Dissertations (1)
- ENGR 310: Technical Writing Final Project (1)
- Graduate School of Architecture & Urban Design Theses & Dissertations (1)
- Mathematics Faculty Research (1)
- Senior Honors Papers / Undergraduate Theses (1)
- WashU Libraries Presentations (1)
- Publication Type
- File Type
Articles 451 - 480 of 1828
Full-Text Articles in Engineering
Influence Of Aerosols During Amine-Based Co2 Capture: Evolution And Control, David Immanuel Arthur Dhanraj
Influence Of Aerosols During Amine-Based Co2 Capture: Evolution And Control, David Immanuel Arthur Dhanraj
McKelvey School of Engineering Graduate Student Theses & Dissertations
Driven by escalating concerns on global warming and climate change, significant efforts have recently been initiated to reduce atmospheric CO2 concentration. Power generation from fossil fuels is the largest source of anthropogenic CO2. The US government has incentivized carbon capture through Section 45Q Tax code, which provides US$50/tCO2 (per metric ton) of qualified carbon dioxide captured and sequestered, as tax credits. The cost for carbon capture, transport and sequestration is US$75 – 150/t CO2, with the state-of-the-art capture technology costing between US$20 – 100/t CO2. Amine based post-combustion CO2 capture is the most economic technology suited for large power plants. …
Mechanisms For Osteoblast And Osteocyte Initiation And Sustainment Of Bone Formation In Young-Adult And Aged Mice, Taylor Lynn Harris
Mechanisms For Osteoblast And Osteocyte Initiation And Sustainment Of Bone Formation In Young-Adult And Aged Mice, Taylor Lynn Harris
McKelvey School of Engineering Graduate Student Theses & Dissertations
The cellular mechanisms for loading-induced bone formation, from osteocyte mechano-sensation to osteoblast-directed bone formation, are not well understood. Elucidating these mechanisms and identifying any processes that are disrupted in aged mice can aide in the development of new anabolic drugs for treating diseases like osteoporosis. This thesis begins by investigating the genes expressed by osteocytes following loading at an early mechanosensitive (4-hr) timepoint, and later at a bone-forming (day 5) timepoint. We demonstrated increases in Ngf and Wnt1 in osteocyte-enriched intracortical bone by laser capture microdissection and microarray analysis. These results were important in demonstrating the presence of Ngf in …
Quantitative Flow Cytometry-Guided Protein Biomarker Characterization For Translational Proteomics And Precision Medicine, Yingye Fang
McKelvey School of Engineering Graduate Student Theses & Dissertations
Angiogenesis, the growth of blood vessels from pre-existing vasculature, is primarily regulated by vascular endothelial growth factor receptors (VEGFRs). Dysregulated angiogenesis is associated with cancers, obesity, and over 70 vascular diseases. Upregulated VEGFR protein expressions in diseased vasculature are promising biomarkers for predicting clinical outcomes, as indicated by non-quantitative immunohistochemical studies in patients with impaired vascularization or tumor angiogenesis. While the quantitative characterization of VEGFRs is critical in identifying biomarkers for anti-angiogenic therapies, VEGFR biomarker development presents two particular challenges: (1) The invasive tissue biopsy needed limits the amount of VEGFR data that can be collected from both normal and …
Attachment Of Fibrous Materials In Nature And Surgical Repair, Ethan Daniel Hoppe
Attachment Of Fibrous Materials In Nature And Surgical Repair, Ethan Daniel Hoppe
McKelvey School of Engineering Graduate Student Theses & Dissertations
Mother nature is the ultimate problem solver. Often by brute force and necessity for survival, the organism that has the best solution to a problem is the one to proliferate. One of these problems for which nature provides an elegant solution is the attachment of dissimilar materials. This dissertation explores the strengthening and relaxation of interactions between soft and hard materials, typically with one of the two being fibrous in character, with the goal of identifying strategies for improving the surgical reattachment of tendon to bone. The work begins with a study of the plant Harpagonella palmeri, which the dissertation …
Understanding And Improving Adult Stem Cells For Cartilage Tissue Engineering, Dakota Katz
Understanding And Improving Adult Stem Cells For Cartilage Tissue Engineering, Dakota Katz
McKelvey School of Engineering Graduate Student Theses & Dissertations
Articular cartilage allows for near frictionless joint movement; however, when damaged the tissue has very little capacity for self-repair. Tissue engineering can be used to both repair damaged articular cartilage and as an in vitro model for joint disease. Commonly used cell sources for tissue engineering are adipose-derived and bone marrow-derived mesenchymal stem cells (ASCs and MSCs) because they can be patient matched, expand rapidly in culture, and have multipotent differentiation potential. However, donor-to-donor variability of differentiation potential can mask the results of in vitro experiments and ASCs and MSCs only retain their multipotency for a limited number of passages. …
Investigation Of Microdroplet Generation, Morphological Evolution, And Applications Under Quasi-Steady And Dynamic States, Li Shan
McKelvey School of Engineering Graduate Student Theses & Dissertations
Microscale droplets are commonly encountered in the fields of materials processing, thermal fluids, and biology. While these droplets are naturally occurring, recent advances in microfabrication have enabled researchers to harness their enhanced transport characteristics for numerous laboratory and industrial applications from controlled chemical synthesis to inkjet printing and thermal management. Smaller droplets have larger specific surface area and a greater perimeter-to-area ratio when resting on a surface (i.e., sessile), which accelerates processes occurring at droplet surfaces like evaporation, chemical reaction, or combustion. The demand for microdroplets with smaller and more uniform sizes has motivated investigation of how such droplets can …
Co-Registered Photoacoustic And Ultrasound Tomographic Imaging Of Human Colorectal And Ovarian Cancer: Light Delivery, System Development, And Clinical Study, Guang Yang
McKelvey School of Engineering Graduate Student Theses & Dissertations
Ovarian cancer remains the deadliest of all the gynecological malignancies. Conventional screening tests, including pelvic examination, transvaginal ultrasound (TVUS), and blood testing for cancer antigen 125 (CA-125), lack sufficient specificity for early ovarian cancer diagnosis. Imaging modalities such as computed tomography (CT), positron emission tomography (PET), and magnetic resonance imaging (MRI) have been used for surgical guidance. However, all of these modalities have limitations in detecting small lesions. Globally, colorectal cancer is the second most commonly diagnosed malignancy and the fourth most common cause of cancer mortality. Accurate staging and post-treatment surveillance of this prevalent disease are critical because treatment …
Algebraic, Computational, And Data-Driven Methods For Control-Theoretic Analysis And Learning Of Ensemble Systems, Wei Miao
McKelvey School of Engineering Graduate Student Theses & Dissertations
In this thesis, we study a class of problems involving a population of dynamical systems under a common control signal, namely, ensemble systems, through both control-theoretic and data-driven perspectives. These problems are stemmed from the growing need to understand and manipulate large collections of dynamical systems in emerging scientific areas such as quantum control, neuroscience, and magnetic resonance imaging. We examine fundamental control-theoretic properties such as ensemble controllability of ensemble systems and ensemble reachability of ensemble states, and propose ensemble control design approaches to devise control signals that steer ensemble systems to desired profiles. We show that these control-theoretic properties …
A Neuromorphic Machine Learning Framework Based On The Growth Transform Dynamical System, Ahana Gangopadhyay
A Neuromorphic Machine Learning Framework Based On The Growth Transform Dynamical System, Ahana Gangopadhyay
McKelvey School of Engineering Graduate Student Theses & Dissertations
As computation increasingly moves from the cloud to the source of data collection, there is a growing demand for specialized machine learning algorithms that can perform learning and inference at the edge in energy and resource-constrained environments. In this regard, we can take inspiration from small biological systems like insect brains that exhibit high energy-efficiency within a small form-factor, and show superior cognitive performance using fewer, coarser neural operations (action potentials or spikes) than the high-precision floating-point operations used in deep learning platforms. Attempts at bridging this gap using neuromorphic hardware has produced silicon brains that are orders of magnitude …
Human Ipsc Tissue-Engineered Cartilage For Disease Modeling Of Skeletal Dysplasia-Causing Trpv4 Mutations, Amanda R. Dicks
Human Ipsc Tissue-Engineered Cartilage For Disease Modeling Of Skeletal Dysplasia-Causing Trpv4 Mutations, Amanda R. Dicks
McKelvey School of Engineering Graduate Student Theses & Dissertations
Cartilage is essential to joint development and function. However, there is a variety of cartilage diseases, ranging from developmental (e.g., skeletal dysplasias) to degenerative (e.g., arthritis), in which treatments and therapeutics are lacking. For example, specific point mutations in the ion channel transient receptor potential vanilloid 4 (TRPV4) prevent proper joint development, leading to mild brachyolmia and severe, neonatally lethal metatropic dysplasia. Tissue-engineered cartilage offers an opportunity to elucidate the underlying mechanisms of these cartilage diseases for the development of treatments. Human induced pluripotent stem cells (hiPSCs) are an improved cell source option for cartilage tissue engineering given their minimal …
Elucidating And Leveraging Dynamics-Function Relationships In Neural Circuits Through Modeling And Optimal Control, Sruti Mallik
Elucidating And Leveraging Dynamics-Function Relationships In Neural Circuits Through Modeling And Optimal Control, Sruti Mallik
McKelvey School of Engineering Graduate Student Theses & Dissertations
A fundamental research question in neuroscience pertains to understanding how neural networks through their activity encode and decode information. In this research, we build on methods from theoretical domains such as control theory, dynamical systems analysis and reinforcement learning to investigate such questions. Our objective is two-fold: first, to use methods from engineering to identify specific objectives that neural circuits might be optimizing through their spatiotemporal activity patterns, and second, to draw motivation from neuroscience to formulate new engineering principles such as synthesis of dynamical networks for decentralized control applications. We specifically take a top-down, optimization driven approach in our …
Long-Term Neural Activity Recorders Using Energy-Based Sensing, Compressive Computation And Data Logging, Darshit Mehta
Long-Term Neural Activity Recorders Using Energy-Based Sensing, Compressive Computation And Data Logging, Darshit Mehta
McKelvey School of Engineering Graduate Student Theses & Dissertations
Insects are ideal candidates for developing bio-robotic systems owing to their ability to thrive in almost any environment. For example, neurons in their exquisite olfactory sensory systems can be tapped to create a sensing platform for standoff chemical monitoring. However, for enabling such cyborg systems, it is vital that the neural activity of a freely behaving organism can be measured for long periods of time. The current state-of-the-art neural recording techniques are power-intensive and they either need batteries, which make them too bulky for insects, or they have to maintain a continuous telemetry link to an external power source which …
Continuous-Time And Complex Growth Transforms For Analog Computing And Optimization, Oindrila Chatterjee
Continuous-Time And Complex Growth Transforms For Analog Computing And Optimization, Oindrila Chatterjee
McKelvey School of Engineering Graduate Student Theses & Dissertations
Analog computing is a promising and practical candidate for solving complex computational problems involving algebraic and differential equations. At the fundamental level, an analog computing framework can be viewed as a dynamical system that evolves following fundamental physical principles, like energy minimization, to solve a computing task. Additionally, conservation laws, such as conservation of charge, energy, or mass, provide a natural way to couple and constrain spatially separated variables. Taking a cue from these observations, in this dissertation, I have explored a novel dynamical system-based computing framework that exploits naturally occurring analog conservation constraints to solve a variety of optimization …
Machine Learning For Analog/Mixed-Signal Integrated Circuit Design Automation, Weidong Cao
Machine Learning For Analog/Mixed-Signal Integrated Circuit Design Automation, Weidong Cao
McKelvey School of Engineering Graduate Student Theses & Dissertations
Analog/mixed-signal (AMS) integrated circuits (ICs) play an essential role in electronic systems by processing analog signals and performing data conversion to bridge the analog physical world and our digital information world.Their ubiquitousness powers diverse applications ranging from smart devices and autonomous cars to crucial infrastructures. Despite such critical importance, conventional design strategies of AMS circuits still follow an expensive and time-consuming manual process and are unable to meet the exponentially-growing productivity demands from industry and satisfy the rapidly-changing design specifications from many emerging applications. Design automation of AMS IC is thus the key to tackling these challenges and has been …
Metal Oxide Nanoparticles Synthesized By Nonthermal Plasma For Electrochemical Applications, Qinyi Chen
Metal Oxide Nanoparticles Synthesized By Nonthermal Plasma For Electrochemical Applications, Qinyi Chen
McKelvey School of Engineering Graduate Student Theses & Dissertations
Metal oxide nanocrystals can be synthesized by colloidal processing and gas-phase processing routes. High-quality nanocrystals produced by colloidal routes typically have surface ligands. Transport of electrons and ions among nanocrystals is hindered by surface ligand, and degrades related electrical performance. Gas-phase processing, for example combustion and thermal plasma, can make free-standing nanocrystals. However, nanocrystals produced by combustion method are heavily agglomerated, which further limits the usage of nanoparticles in electrochemical applications. Nonthermal plasma, as an emerging attractive route, has been initially focused on the synthesis of elemental group IV materials, and extended to include the synthesis of multicomponent materials, such …
Developing An Induction Protocol And Generating A Selectable Cell Line To Obtain High-Purity V0v Interneuron Cultures From Mouse Embryonic Stem Cells, Jennifer Pardieck
Developing An Induction Protocol And Generating A Selectable Cell Line To Obtain High-Purity V0v Interneuron Cultures From Mouse Embryonic Stem Cells, Jennifer Pardieck
McKelvey School of Engineering Graduate Student Theses & Dissertations
Spinal cord injury (SCI) leads to loss of motor function among other devastating impairments, with patients having a grim outlook on recovery as currently there are no widely efficacious treatment methods. The native regenerative capacity of the neurons in the central nervous system is limited, thus recovery of the axonal connections between the different components of the complex circuitry that enables motor function is minimal after SCI. Investigators aiming to improve motor function post-SCI approach the problem from different angles, including removing inhibitory elements to axonal regeneration and supporting regeneration through implantation of biomaterials, exposure to pro-regenerative factors, and transplantation …
Effect Of Ultrasound Pressure On The Distribution Of Bovine Serum Albumin Delivered By Focused Ultrasound-Mediated Blood-Brain Barrier Opening In Cleared Mouse Brains, Yajie Liu
McKelvey School of Engineering Graduate Student Theses & Dissertations
Most common diagnosis and therapeutic methods have low effectiveness when used on brain diseases. The key obstacle is that the blood-brain barrier (BBB) prevents most drugs from entering the brain. Some strategies have been developed to improve the efficiency of drug delivery crossing BBB. Among all these strategies, focused ultrasound-mediated BBB opening (FUS-BBB Opening) stands out since it is noninvasive and can be located to the target area. Detailed studies are required on the distribution of drugs delivered by FUS-BBB opening and the effects of FUS parameters on the distribution. This thesis proposes a pipeline involving tissue clearing and lightsheet …
Jme 4110: Water Reservoir Energy Storage, Kristen Grace, Supapan Thammanok, Leticia Zoratto Lunge
Jme 4110: Water Reservoir Energy Storage, Kristen Grace, Supapan Thammanok, Leticia Zoratto Lunge
Washington University / UMSL Mechanical Engineering Design Project JME 4110
The goal of this project was to create a demonstrative model of a Pumped Hydroelectric Storage System (PHS). A PHS stores energy by draining water from an upper reservoir through a turbine to generate electricity. This water is then pumped from a lower reservoir to the original upper reservoir to restart the system cycle. Using existing PHS applications, a small-scale prototype was built using an aquarium pump and mini turbine, along with PVC and plastic containers. When water is run through the system, a multimeter can be used to show the power output of the system. In addition to functionality, …
Jme 4110: Extended Hole Punch, Lucas Mennemeier, Jason Zobrist
Jme 4110: Extended Hole Punch, Lucas Mennemeier, Jason Zobrist
Washington University / UMSL Mechanical Engineering Design Project JME 4110
DESIGN AND DEVELOP A WORKING PROTOTYPE OF A SINGLE-HAND, SINGLE-HOLE PUNCH WITH AN EXTENDED THROAT THAT WILL BE ABLE TO REACH FURTHER INTO A SHEET OF PAPER THAN A STANDARD SINGLE-HOLE PUNCH. THIS CAN BE A PUNCH ATTACHMENT OR A STAND-ALONE DEVICE. EITHER DESIGN CHOSEN WILL ALSO BE ABLE TO BE CONVERTED INTO AN EMBOSSING TOOL FOR EMBOSSING SEALS INTO DOCUMENTS
Jme 4110: Drag Line Wind Energy Generator, David Gengenbacher, Mouhamadou Bamba Kane, Brian Bakula
Jme 4110: Drag Line Wind Energy Generator, David Gengenbacher, Mouhamadou Bamba Kane, Brian Bakula
Washington University / UMSL Mechanical Engineering Design Project JME 4110
The goal of this project was to develop a prototype capable of power a 10 watt light bulb using only the energy from wind.
Jme 4110: Mass Production Of Multi-Sphere Approximated Parts, Christopher Hartley, Chad Gorski, Matthew Brady
Jme 4110: Mass Production Of Multi-Sphere Approximated Parts, Christopher Hartley, Chad Gorski, Matthew Brady
Washington University / UMSL Mechanical Engineering Design Project JME 4110
For this project, the team needs to design a mass production manufacturing process. The product of the process will be parts that represent, as closely as possible, approximations of parts developed using Discrete Element Modelling (DEM) software. Since the software uses spheres to efficiently model complex parts, the manufacturing process must also utilize spheres as the building blocks for the approximated parts. As a proof of concept, this project uses river pebbles as the parts being approximated by the software and modeled by the multi-sphere production process.
Jme 4110: Lego-Lock Block, Diana Engelhardt, Dmitriy Kornyukhov, Chelsea Borders
Jme 4110: Lego-Lock Block, Diana Engelhardt, Dmitriy Kornyukhov, Chelsea Borders
Washington University / UMSL Mechanical Engineering Design Project JME 4110
Design a building block device that assembles similarly to a LEGO brick, but also has the capability to lock an assembly state between two bricks. Scale (size) not critical for this proof of concept (standard LEGO blocks may be too small for this first attempt).
Jme 4110: Vibratory Parts Feeder, Adin Stambolic, Patrick Edward Vastola, Noah Herrin
Jme 4110: Vibratory Parts Feeder, Adin Stambolic, Patrick Edward Vastola, Noah Herrin
Washington University / UMSL Mechanical Engineering Design Project JME 4110
Design an inexpensive, modular and tunable vibratory feeder to move different size particles from one place to another. The feeder can change according to the shape and size of the particles while in use.
Jme 4110: Data Cooling Center, Christopher Schmidt, Brennan Fogarty, Suruchi Aryal
Jme 4110: Data Cooling Center, Christopher Schmidt, Brennan Fogarty, Suruchi Aryal
Washington University / UMSL Mechanical Engineering Design Project JME 4110
The group will design a way to efficiently move heat within a framework of using dielectric fluid as a heat transfer medium. The heat will be removed from a data center or a simulated data center. The group will evaluate the best possible design with the understanding that cooling data centers generate a large amount of heat. The design will also be constructed in such a manner that it can be scaled up to a larger size. Dielectric fluid will be used due to its efficiency compared to that of an air cooled system or a compressed refrigerant system. Account …
Jme 4110: Portable Car Wash, Coleman Kaufman, Sara Wilson
Jme 4110: Portable Car Wash, Coleman Kaufman, Sara Wilson
Washington University / UMSL Mechanical Engineering Design Project JME 4110
We met with our target customer to understand their needs for the ideal portable car wash. Their key needs included use by any sized passenger vehicle, washing the car as good as a community car wash, and must be portable. We started by researching current products that could match these needs, and then we sketched four of our own designs to match our customer’s needs. With these design sketches we used a metrics tool to determine which design was the best. From there we did engineering calculations to analyze the trouble spots in our design, such as the water pressure …
Investigation Of Buoyancy-Driven Convection In Non-Axisymmetric Evaporating Droplets Confined On Hollow Micropillars, De Dong
McKelvey School of Engineering Graduate Student Theses & Dissertations
Sessile droplet evaporation is ubiquitous in natural and industrial processes, including mammal perspiration, transpiration by plants, and industrial spray cooling. Although extensive evaporation-related research has examined convection effects inside droplets, most studies have been based on hemispherical droplets. Hence the underlying mechanisms of evaporation from non-axisymmetric droplets are not well understood. Here, we investigate the mechanisms of evaporation from non-axisymmetric droplets with sizes ranging from 100 μm to 8 mm. An experimentally validated finite element model simulates different transport mechanisms during evaporation. For non-axisymmetric droplets larger than 4 mm, we observe a non-uniform recirculation pattern in the azimuthal direction, leading …
Optimizing Biomaterials For Ipsc-Derived Engineered Heart Tissue Models, Huanzhu Jiang
Optimizing Biomaterials For Ipsc-Derived Engineered Heart Tissue Models, Huanzhu Jiang
McKelvey School of Engineering Graduate Student Theses & Dissertations
Hypertrophic Cardiomyopathy (HCM) is the most common inherited cardiovascular disease, where environmental forces and genetic mutations have been heavily linked to heart failure or sudden cardiac death in young people. However, how genetic mutations and environmental forces contribute to HCM phenotype are not well defined. To study the mechanisms and investigate the potential treatments of this disease, it is important to build a model to simulate mechanical loading and recreate micro-environment of cardiomyocytes. In vivo, surgical maneuvers like transaortic constriction allow the study of overload, but these approaches are low throughput and it is difficult to precisely control overload. …
Development Of A New Transitional Flow Model Integrating The One-Equation Wray-Agarwal Turbulence Model With An Algebraic Intermittency Transport Term, Yan Xue
McKelvey School of Engineering Graduate Student Theses & Dissertations
No abstract provided.
Injectable Ct/Mri Contrast Agent For Gastrointestinal Tumor Tracking, Luna Zhang
Injectable Ct/Mri Contrast Agent For Gastrointestinal Tumor Tracking, Luna Zhang
McKelvey School of Engineering Graduate Student Theses & Dissertations
Gastrointestinal cancers remain to be of the most common and deadly cancers worldwide. Early detection and treatments are crucial for reducing mortality and improving patient outcome. Radiation therapy is a non-invasive localized tumor treatment method, and utilizes radiation to kill the cancerous cells and shrink tumors at specific sites. Precise localization at the target tumor site is therefore important before radiation therapy, especially for gastrointestinal tumor sites located in the moving bowel. Currently, invasive endoscopies along with ink tattoos are used for identifying tumor location, which often require sedation and bring much discomfort. Imaging tests, including CT and MRI, play …
Improving Additional Adversarial Robustness For Classification, Michael Guo
Improving Additional Adversarial Robustness For Classification, Michael Guo
McKelvey School of Engineering Graduate Student Theses & Dissertations
Although neural networks have achieved remarkable success on classification, adversarial robustness is still a significant concern. There are now a series of approaches for designing adversarial examples and methods to defending against them. This paper consists of two projects. In our first work, we propose an approach by leveraging cognitive salience to enhance additional robustness on top of these methods. Specifically, for image classification, we split an image into the foreground (salient region) and background (the rest) and allow significantly larger adversarial perturbations in the background to produce stronger attacks. Furthermore, we show that adversarial training with dual-perturbation attacks yield …