Computer Sciences Commons^™

Investigating Intervention Components And Exploring States Of Receptivity For A Smartphone App To Promote Physical Activity: Study Protocol Of The Ally Micro-Randomized Trial, Jan-Niklas Kramer, Florian Künzler, Varun Mishra, Bastien Presset, David Kotz, Shawna Smith, Urte Scholz, Tobias Kowatsch

Dartmouth Scholarship

Background: Smartphones enable the implementation of just-in-time adaptive interventions (JITAIs) that tailor the delivery of health interventions over time to user- and time-varying context characteristics. Ideally, JITAIs include effective intervention components, and delivery tailoring is based on effective moderators of intervention effects. Using machine learning techniques to infer each user’s context from smartphone sensor data is a promising approach to further enhance tailoring.

Objective: The primary objective of this study is to quantify main effects, interactions, and moderators of 3 intervention components of a smartphone-based intervention for physical activity. The secondary objective is the exploration of participants’ states of receptivity, …

Spice: Secure Proximity-Based Infrastructure For Close Encounters, Aarathi Prasad, Xiaohui Liang, David Kotz

Dartmouth Scholarship

We present a crowdsourcing system that extends the capabilities of location-based applications and allows users to connect and exchange information with users in spatial and temporal proximity. We define this incident of spatio-temporal proximity as a \em close encounter. Typically, location-based application users store their information on a server, and trust the server to provide access only to authorized users, not misuse the data or disclose their location history. Our system, called SPICE, addresses these privacy issues by leveraging Wi-Fi access points to connect users and encrypt their information before it is exchanged, so only users in close encounters have …

A Statistical Prior For Photo Forensics: Object Removal, Wei Fan, Hany Farid

Computer Science Technical Reports

If we consider photo forensics within a Bayesian framework, then the probability that an image has been manipulated given the results of a forensic test can be expressed as a product of a likelihood term (the probability of a forensic test detecting manipulation given that an image was manipulated) and a prior term (the probability that an image was manipulated). Despite the success of many forensic techniques, the incorporation of a statistical prior has not been previously considered. We describe a framework for incorporating statistical priors into any forensic analysis and specifically address the problem of quantifying the probability that …

A Digital Practice Tool For Chemical Resonance, Kimball Jaclyn

Dartmouth College Master’s Theses

Digital practice tools support online learning in math, language, computer science, and other subjects, but practice with problems whose answers are not well represented by text or quantities is underrepresented in the digital learning ecosystem beyond multiple-choice questions. This thesis project explored an alternative to multiple choice practice problems in organic chemistry that does not rely on a molecule drawing interface. This project included development and evaluation of a proof-of-concept digital practice tool for chemical resonance problems. Results of a utility study strongly suggest that the practice tool could fill a learning niche within organic chemistry practice as part of …

Shareabel: Secure Sharing Of Mhealth Data Through Cryptographically-Enforced Access Control, Emily Greene

Dartmouth College Undergraduate Theses

Owners of mobile-health apps and devices often want to share their mHealth data with others, such as physicians, therapists, coaches, and caregivers. For privacy reasons, however, they typically want to share a limited subset of their information with each recipient according to their preferences. In this paper, we introduce ShareABEL, a scalable, usable, and practical system that allows mHealth-data owners to specify access-control policies and to cryptographically enforce those policies so that only parties with the proper corresponding permissions are able to decrypt data. The design (and prototype implementation) of this system makes three contributions: (1) it applies cryptographically-enforced access-control …

Toward Accurate And Efficient Feature Selection For Speaker Recognition On Wearables, Rui Liu, Reza Rawassizadeh, David Kotz

Dartmouth Scholarship

Due to the user-interface limitations of wearable devices, voice-based interfaces are becoming more common; speaker recognition may then address the authentication requirements of wearable applications. Wearable devices have small form factor, limited energy budget and limited computational capacity. In this paper, we examine the challenge of computing speaker recognition on small wearable platforms, and specifically, reducing resource use (energy use, response time) by trimming the input through careful feature selections. For our experiments, we analyze four different feature-selection algorithms and three different feature sets for speaker identification and speaker verification. Our results show that Principal Component Analysis (PCA) with frequency-domain …

Opencollab: A Blockchain Based Protocol To Incentivize Open Source Software Development, Yondon Fu

Dartmouth College Undergraduate Theses

Open source software is one of the fundamental building blocks of today's technology dependent society and is relied upon by parties ranging from large technology corporations to individual hobbyist developers. The open question left for technologists is how to make open source software projects more sustainable. The rise of decentralized networks of self-organizing, self-coordinating users incentivized by valuable cryptographic tokens enabled by Ethereum smart contracts creates the possibility of a system with embedded economics for open source software development that aligns the incentives of all parties. We present two contributions that can serve as building blocks for a potentially better …

Novenanetwork: A Catholic Social Media Ios Application For Praying Novenas As A Community, Marissa Le Coz

Dartmouth College Undergraduate Theses

Recent years have seen the rise of a Catholic presence on the Internet and on social media in particular. At the same time, there has been a movement among young Catholics to rediscover and revitalize the traditions of the Church. Novenas, an ancient form of spirituality involving praying for nine consecutive days for some special purpose, are one such tradition. NovenaNetwork, which debuted in the App Store on May 23, 2017 and acquired over 50 users in three countries within its first six days, is an iOS social media application for praying novenas with others. While multiple apps and websites …

Chinese Font Style Transfer With Neural Network, Xue Hanyu

Dartmouth College Master’s Theses

Font design is an important area in digital art. However, designers have to design character one by one manually. At the same time, Chinese contains more than 20,000 characters. Chinese offical dataset GB 18030-2000 has 27,533 characters. ZhongHuaZiHai, an official Chinese dictionary, contains 85,568 characters. And JinXiWenZiJing, an dataset published by AINet company, includes about 160,000 chinese characters. Thus Chinese font design is a hard task. In the paper, we introduce a method to help designers finish the process faster. With the method, designers only need to design a small set of Chinese characters. Other characters will be generated automatically. …

Accuracy And Racial Biases Of Recidivism Prediction Instruments, Julia J. Dressel

Dartmouth College Undergraduate Theses

Algorithms have recently become prevalent in the criminal justice system. Tools known as recidivism prediction instruments (RPIs) are being used all over the country to assess the likelihood that a criminal defendant will reoffend at some point in the future. In June of 2016, researchers at ProPublica published an analysis claiming an RPI called COMPAS was biased against black defendants. This claim sparked a nation-wide debate as to how fairness of an algorithm should be measured, and exposed the many ways that algorithms are not necessarily fair. Algorithms are used in the criminal justice system because they are regarded as …

Assemble.Live: Designing For Schisms In Large Groups In Audio/Video Calls, Benjamin P. Packer

Dartmouth College Undergraduate Theses

Although new communication technologies have compressed the space and latency between participants, leading to new forms of computer mediated interaction that scale with the number of participants [Klein, 1999], there still exist no audio/video calling solutions that can accommodate the type of group conversation that takes place in a group of four or more. Groups of this size frequently schism, forming two or more sub-conversations with their own independently operating turn taking systems [Egbert, 1997]. This paper proposes that traditional audio/video calling fails to accommodate schisms because a) there is no way to signal intended recipiency, b) there exists only …

Using Computational Models To Understand Asd Facial Expression Recognition Patterns, Irene L. Feng

Dartmouth College Undergraduate Theses

Recent advances in computer vision have led to interest in studying how computer vision can simulate our own perception to better understand the intricacies of human neurobiology. Researchers have made strides in computer vision to imitate many facets of human perception, such as object detection, character recognition, and face identification. However, there have been fewer studies that try to model atypical human perception. My thesis focuses specifically on individuals with Autism Spectrum Disorder (ASD) and their deficit in the facial expression recognition (FER) task. I built multiple computer vision models using hand-crafted features and also convolutional neural network architectures to …

Improving Elementary Math Learning Through Ipad Games, Kaya M. Thomas

Dartmouth College Undergraduate Theses

Mathematics has proven to be challenging to many from a very young age. Young students are influenced by their teachers on how to feel about math and how well they can perform. Currently many methods of teaching mathematics do not encourage learning, but instead promote memorization which has been shown to increase students' anxiety about math. Math anxiety affects student performance as well as their ability to understand the material. Fractions are one of the most difficult concepts for young students to learn. Various techniques have been created in order to better instruct students on how to understand fractions. More …

Scene Classification From Degraded Images: Comparing Human And Computer Vision Performance, Tim M. Tadros

Dartmouth College Undergraduate Theses

People can recognize the context of a scene with just a brief glance. Visual information such as color, objects and their properties, and texture are all important in correctly determining the type of scene (e.g. indoors versus outdoors). Although these properties are all useful, it is unclear which features of an image play a more important role in the task of scene recognition. To this aim, we compare and contrast a state-of-the-art neural network and GIST model with human performance on the task of classifying images as indoors or outdoors. We analyze the impact of image manipulations, such as blurring …

A Survey Of Trustworthy Computing On Mobile & Wearable Systems, Travis Peters

Computer Science Technical Reports

Mobile and wearable systems have generated unprecedented interest in recent years, particularly in the domain of mobile health (mHealth) where carried or worn devices are used to collect health-related information about the observed person. Much of the information - whether physiological, behavioral, or social - collected by mHealth systems is sensitive and highly personal; it follows that mHealth systems should, at the very least, be deployed with mechanisms suitable for ensuring confidentiality of the data it collects. Additional properties - such as integrity of the data, source authentication of data, and data freshness - are also desirable to address other …

Activityaware: Wearable System For Real-Time Physical Activity Monitoring Among The Elderly, George G. Boateng

Dartmouth College Master’s Theses

Physical activity helps reduce the risk of cardiovascular disease, hypertension and obesity. The ability to monitor a person’s daily activity level can inform self-management of physical activity and related interventions. For older adults with obesity, the importance of regular, physical activity is critical to reduce the risk of long-term disability. In this work, we present ActivityAware, an application on the Amulet wrist-worn device that monitors the daily activity levels (low, moderate and vigorous) of older adults in real-time. The app continuously collects acceleration data on the Amulet, classifies the current activity level, updates the day’s accumulated time spent at that …

Cryptographic Transfer Of Sensor Data From The Amulet To A Smartphone, David B. Harmon

Dartmouth College Undergraduate Theses

The authenticity, confidentiality, and integrity of data streams from wearable healthcare devices are critical to patients, researchers, physicians, and others who depend on this data to measure the effectiveness of treatment plans and clinical trials. Many forms of mHealth data are highly sensitive; in the hands of unintended parties such data may reveal indicators of a patient's disorder, disability, or identity. Furthermore, if a malicious party tampers with the data, it can affect the diagnosis or treatment of patients, or the results of a research study. Although existing network protocols leverage encryption for confidentiality and integrity, network-level encryption does not …

A Hololens Application To Aid People Who Are Visually Impaired In Navigation Tasks, Jonathan L. Huang

Dartmouth College Undergraduate Theses

Day-to-day activities such as navigation and reading can be particularly challenging for people with visual impairments. Reading text on signs may be especially difficult for people who are visually impaired because signs have variable color, contrast, and size. Indoors, signage may include office, classroom, restroom, and fire evacuation signs. Outdoors, they may include street signs, bus numbers, and store signs. Depending on the level of visual impairment, just identifying where signs exist can be a challenge. Using Microsoft's HoloLens, an augmented reality device, I designed and implemented the TextSpotting application that helps those with low vision identify and read indoor …

Lighttouch: Securely Connecting Wearables To Ambient Displays With User Intent, Xiaohui Liang, Tianlong Yun, Ronald Peterson, David Kotz

Dartmouth Scholarship

Wearables are small and have limited user interfaces, so they often wirelessly interface with a personal smartphone/computer to relay information from the wearable for display or other interactions. In this paper, we envision a new method, LightTouch, by which a wearable can establish a secure connection to an ambient display, such as a television or a computer monitor, while ensuring the user's intention to connect to the display. LightTouch uses standard RF methods (like Bluetooth) for communicating the data to display, securely bootstrapped via the visible-light communication (the brightness channel) from the display to the low-cost, low-power, ambient light sensor …

Quantitative Criticism Of Literary Relationships, Joseph P. Dexter, Theodore Katz, Nilesh Tripuraneni, Tathagata Dasgupta, Ajay Kannan, James Brofos, Jorge A. Bonilla Lopez, Lea Schroeder

Dartmouth Scholarship

Authors often convey meaning by referring to or imitating prior works of literature, a process that creates complex networks of literary relationships (“intertextuality”) and contributes to cultural evolution. In this paper, we use techniques from stylometry and machine learning to address subjective literary critical questions about Latin literature, a corpus marked by an extraordinary concentration of intertextuality. Our work, which we term “quantitative criticism,” focuses on case studies involving two influential Roman authors, the playwright Seneca and the historian Livy. We find that four plays related to but distinct from Seneca’s main writings are differentiated from the rest of the …

Dense Gray Codes In Mixed Radices, Jessica C. Fan

Dartmouth College Undergraduate Theses

The standard binary reflected Gray code describes a sequence of integers 0 to n-1, where n is a power of 2, such that the binary representation of each integer in the sequence differs from the binary representation of the preceding integer in exactly one bit. In September 2016, we presented two methods to compute binary dense Gray codes, which extend the possible values of n to the set of all positive integers while preserving both the Gray-code property such that only one bit changes between each pair of consecutive binary numbers, and the density property such that the sequence contains …

Activityaware: An App For Real-Time Daily Activity Level Monitoring On The Amulet Wrist-Worn Device, George Boateng, Ryan Halter, John A. Batsis, David Kotz

Dartmouth Scholarship

Physical activity helps reduce the risk of cardiovascular disease, hypertension and obesity. The ability to monitor a person's daily activity level can inform self-management of physical activity and related interventions. For older adults with obesity, the importance of regular, physical activity is critical to reduce the risk of long-term disability. In this work, we present ActivityAware, an application on the Amulet wrist-worn device that measures daily activity levels (sedentary, moderate and vigorous) of individuals, continuously and in real-time. The app implements an activity-level detection model, continuously collects acceleration data on the Amulet, classifies the current activity level, updates the day's …

The Amulet Wearable Platform: Demo Abstract, Josiah Hester, Travis Peters, Tianlong Yun, Ronald Peterson, Joseph Skinner, Bhargav Golla, Kevin Storer, Steven Hearndon, Sarah Lord, Ryan Halter, David Kotz, Jacob Sorber

Dartmouth Scholarship

In this demonstration we present the Amulet Platform; a hardware and software platform for developing energy- and resource-efficient applications on multi-application wearable devices. This platform, which includes the Amulet Firmware Toolchain, the Amulet Runtime, the ARP-View graphical tool, and open reference hardware, efficiently protects applications from each other without MMU support, allows developers to interactively explore how their implementation decisions impact battery life without the need for hardware modeling and additional software development, and represents a new approach to developing long-lived wearable applications. We envision the Amulet Platform enabling long-duration experiments on human subjects in a wide variety of studies.

Representations And Models For Large-Scale Video Understanding, Du Le Hong Tran

Dartmouth College Ph.D Dissertations

In this thesis, we investigate different representations and models for large-scale video understanding. These methods include a mid-level representation for action recognition, a deep-learned representation for video analysis, a generic convolutional network architecture for video voxel prediction, and a new high-level task and benchmark of video comprehension. First, we present EXMOVES, a mid-level representation for scalable action recognition. The entries in EXMOVES representation are the calibrated outputs of a set of movement classifiers over spatial-temporal volumes of the input video. Each movement classifier is a simple exemplar-SVM trained on low-level features. Our EXMOVES requires a minimal amount of supervision while …

Monte Carlo Convergence Analysis For Anisotropic Sampling Power Spectra, Gurprit Singh, Wojciech Jarosz

Computer Science Technical Reports

Traditional Monte Carlo (MC) integration methods use point samples to numerically approximate the underlying integral. This approximation introduces variance in the integrated result, and this error can depend critically on the sampling patterns used during integration. Most of the well known samplers used for MC integration in graphics, e.g. jitter, Latin hypercube (n-rooks), multi-jitter, are anisotropic in nature. However, there are currently no tools available to analyze the impact of such anisotropic samplers on the variance convergence behavior of Monte Carlo integration. In this work, we propose a mathematical tool in the Fourier domain that allows analyzing the variance, and …

Artificial Intelligence And Amikacin Exposures Predictive Of Outcomes In Multidrug-Resistant Tuberculosis Patients, Chawangwa Modongo, Jotam G. Pasipanodya, Shashikant Srivastava, Nicola Zetola, Scott Williams, Giorgio Sirugo, Tawanda Gumbo

Dartmouth Scholarship

Aminoglycosides such as amikacin continue to be part of the backbone of treatment of multidrug-resistant tuberculosis (MDR- TB). We measured amikacin concentrations in 28 MDR-TB patients in Botswana receiving amikacin therapy together with oral levofloxacin, ethionamide, cycloserine, and pyrazinamide and calculated areas under the concentration-time curves from 0 to 24 h (AUC0 –24). The patients were followed monthly for sputum culture conversion based on liquid cultures. The median duration of amikacin therapy was 184 (range, 28 to 866) days, at a median dose of 17.30 (range 11.11 to 19.23) mg/kg. Only 11 (39%) pa- tients had sputum culture conversion during …

The Darklight Rises: Visible Light Communication In The Dark, Zhao Tian, Kevin Wright, Xia Zhou

Computer Science Technical Reports

Visible Light Communication (VLC) emerges as a new wireless communication technology with appealing benefits not present in radio communication. However, current VLC designs commonly require LED lights to emit shining light beams, which greatly limits the applicable scenarios of VLC (e.g., in a sunny day when indoor lighting is not needed). It also entails high energy overhead and unpleasant visual experiences for mobile devices to transmit data using VLC. We design and develop DarkLight, a new VLC primitive that allows light-based communication to be sustained even when LEDs emit extremely-low luminance. The key idea is to encode data into ultra-short, …

Demo: Wanda, Securely Introducing Mobile Devices, Timothy J. Pierson, Xiaohui Liang, Ronald Peterson, David Kotz

Dartmouth Scholarship

Nearly every setting is increasingly populated with wireless and mobile devices – whether appliances in a home, medical devices in a health clinic, sensors in an industrial setting, or devices in an office or school. There are three fundamental operations when bringing a new device into any of these settings: (1) to configure the device to join the wireless local-area network, (2) to partner the device with other nearby devices so they can work together, and (3) to configure the device so it connects to the relevant individual or organizational account in the cloud. The challenge is to accomplish all …

Security And Privacy Analysis Of Medical Wearables, Brian M. Chalif

Dartmouth College Undergraduate Theses

The release of Google glass and the Apple and Samsung smart watches in the past two years pushed wearables to the forefront of technology. The realm of medical wearables will specifically see a huge growth as wearables become more common place. There are documented cases of security and privacy breaches in the five main potential breaching areas: wearable device itself, Bluetooth communication, smartphone or personal computer app, Wi-Fi data exchange, cloud storage. Privacy policies for individual wearables are not always in the best interest of the individual and government regulations on wearables security does not always fully vet wearables. Interviews …

Bae: Before Anyone Else; The Answer To Mobile Dating For The African Diaspora, Jordan S. Kunzika

Dartmouth College Undergraduate Theses

As online dating becomes increasingly popular, millions of people across the country utilize various dating services that allow them to meet, chat, date, and develop meaningful relationships with people they would probably not have met otherwise. While mass market dating apps work well for the majority of online daters, and while many niche dating apps have been created to target a number of specific demographics such as the Jewish population, Southeast Asian population, farmers, homosexuals, etc., the African-American market has been largely underserved. African-Americans have the worst experience on mass market dating apps due to negative racial bias, and before …