October 23rd,
24th and 25th, 2015


Presented by
Faculty & Students in Carnegie Mellon's

School of Computer Science

2015 Our Sponsors:




Computational Biology Dept.

Computer Science Dept.

Human-Computer Interaction Institute

Institute for Software Research

Machine Learning Department

Robotics Institute



2015 Research Team Leaders and Projects

Team Leaders and Project Descriptions

alanBAlan Black
Language Technologies Institute
Carnegie Mellon

Mimic: An Automatic System to Mimic a Human Speech

This project will involve the construction of a system that will take examples of a particular person's speech, and build a speech synthesizer that will mimic their speech pattern so that other speech said by the system will sound like that person.

Members of the project will build synthesized versions of their own voices, and use them to speak, and sing.  We will record a person's voice and use existing tools to build a synthetic version that can say anything which still modeling the
original persons accent, style etc.

emmaEmma Brunskill
Assistant Professor
Computer Science Dept.
Carnegie Mellon

Towards Artificial Agents that Learn to Teach
MOOCs are helping transform access to high quality education but there exists an enormous opportunity to revolutionize the quality of education through machine learning and artificial intelligence.

In this project we will use machine learning to help model student behavior and improve the automated instructional decisions made for an intelligent tutoring system for teaching histograms (a commonly misunderstood topic) to students.

Lorrie Cranor
Associate Professor
Institute for Software Research
Carnegie Mellon
Abby (Abigail) Marsh
Graduate Student
Institute for Software Research
Carnegie Mellon

Understanding Facebook Users’ Privacy-Related Attitudes and Behaviors
As has been visible in recent news, making personal information available on Facebook can negatively impact a user’s offline life. In a recently-publicized example, a teenage girl accidently posted an invitation to her birthday party publicly and received over 21,000 responses. Facebook information being used in unexpected ways has also led to a variety of other negative consequences, such as job loss and loss of health insurance. Despite these incidents, many people still choose to post photos, personal information, and messages online using Facebook. Understanding why users post information online, despite the potential dangers, raises a variety of interesting questions: Do Facebook users believe that they are appropriately using Facebook privacy settings to protect the information they post online? Do Facebook users have a lower privacy threshold than those who choose not to use Facebook? Do Facebook users adopt privacy strategies, successful or not, outside of Facebook’s settings to protect their information? For this project participants will develop a survey to address some of these questions and explore Facebook users’ and nonusers’ privacy preferences. We will gather and analyze actual user data using Amazon’s Mechanical Turk marketplace, allowing participants to draw conclusions around these important issues. 

rogerRoger Dannenberg
Professor of Computer Science, Art and Music
Computer Science Dept. Carnegie Mellon

Anders Øland
Graduate Student
Computer Science Dept.
Carnegie Mellon

gusGus Xia
Ph.D. Candidate
Machine Learning Dept.
Carnegie Mellon

Automatic Generation of Drum Tracks
What makes a good beat? How do drummers rock? Or swing? We can synthesize "perfect" drumming according to standard drumming patterns, but the results are disappointing. In this project, we will analyze real drumming data and try to discover what makes real drumming so much more interesting and musical than naive computer-generated drumming. Our goal is to make a computer drummer that competes with a human. We will begin by considering representations and models of drumming. Using statistics and machine learning techniques, we will fit these models to real drumming data and test how well these models explain the data. Using these models, we can synthesize new drumming patterns. We will discuss evaluation techniques including human evaluation of synthesized drumming.

Bonnie Holub
Director, Risk and Compliance Systems & Analytics, PwC
Minneapolis, MN

Data Analytics and Visualization
Data Analytics and Visualization is the search for meaning in vast repositories of data that swamp scientists, engineers, planners and business people every day. Finding the story in the data and communicating it compellingly requires a skillset that will be required over and over in the future.  The Harvard Business Review calls “Data Scientist” the sexiest job of the 21st century.  The McKinsey Global Institute projects a shortfall of 140,000 to 190,000 data scientists in the United States (and a gap of 1.5 million additional manager and analysts to consume the results) by 2018.

georgeGeorge Kantor
Senior Systems Scientists
Robotics Institute
Carnegie Mellon

Merritt Jenkins
Graduate Student
Robotics Institute
Carnegie Mellon

Robots and Cameras for Automated Plant Phenotyping 
This project will investigate the problem of using robotics, cameras, and 3D modeling technology to better understand plant growth.  This is part of an emerging research area called “high throughput phenotyping”, where automated technologies are used to gather plant information that can be used both for high efficiency plant production and for breeding plants with desirable characteristics.  After reviewing the state of the art in this area, we will brainstorm methods for the automated collection and processing of visual data.  The goal is to build a prototype robotic system that can collect plant images, process the images into high-resolution 3D models, and automatically extract useful phenotype (e.g., stem width or leaf shape) information from the models.

Robert Kraut
Human Computer Interaction Institute
Carnegie Mellon
Carolyn Rosé
Associate Professor Language Technologies Institute
Carnegie Mellon

yichiaYi-Chia Wang
Graduate Student
Language Technologies Institute
Carnegie Mellon


Does this post make me look good on Facebook?
On sites like Facebook, users manage their profiles, likes, and status updates to present a certain image to the world--or at least to their Facebook friends. Websites like Facebook and Twitter are known as "social networking sites" (SNS) for good reason. As social spaces, users often make the same social calculations that they do at job interviews or dates on how to present themselves to look good. People want to make a good impression online, but we know little about what kind of content makes a good impression in users’ minds. In this project, students will use natural language processing and machine learning techniques to examine how Facebook users use language to present themselves positively on the site.

James Morris
Professor and Former Dean of SCS
Human Computer Interaction Institute
Carnegie Mellon

Invent a Mobile Service 
Smart phones are today's computing platform, especially in the developing world. Our team will develop an idea for a new mobile service. Some examples:

Grapevine (invented by a previous OurCS team): Support responsible exchange of opinions about people. 
Favor Net (invented by a previous OurCS team): Facilitate requesting and performing favors on a college campus.
Family Memory Book: Billions of photos are being taken of children every day. How will their parents, the children grown, and their descendants enjoy them?
Product Finder: Snap a picture of an item you like, from TV or real life, and  be told where it came from.
The team will decide on a service and develop usage scenarios.

rockyRocky Medure
User Experience Designer
Google, Pittsburgh

Timothy James
Partner Technology Manager
Google, Pittsburgh


Gopi Flaherty
Software Engineer in Test
Google, Pittsburgh

Daniel Klein
Software and Site Reliability Engineer
Google, Pittsburgh


Exploring the Intelligent Fabric of our World
Computers have become the modern backbone of communication, which creates a powerful fabric for interacting with the environment and objects around us. We'll explore the space of microcontrollers, sensors, and peripherals that enable us to become creators of a future intelligent world. Students will use this fabric to make a prototype that allows them to communicate with and control an object in their world.

Examples include wearable computing, telepresence systems, and remote monitoring.

danielDaniel Mosse
Professor and Chair of Computer Science Department
University of Pittsburgh

grishmaGrishma Jena
OurCS Alumna
Graduate Student
Computer Science Department
University of Pennsylvania

Energy Storage and Distribution in Renewable-energy Powered Microgrids

Research challenges being addressed:

  1. Local energy stability and resiliency: create algorithm to maximize energy harvesting and sharing in microgrids
  2. Efficiently storing energy: minimizing the amount of energy storage and using different energy storage technologies

Abstract: In a microgrid, a set of households can share energy harnessed locally from renewable sources.  Renewable power (e.g., wind power) is well known for being intermittent.  That intermittency might not synchronize with the microgrid’s power demand.  That problem can be softened by storing renewable energy.  However, the cost of storing energy in the form of electricity is viewed as a barrier to using renewable power.  This research seeks to reduce that cost by scheduling energy to be stored in several forms, such as chemical batteries (as usual) but also new dedicated energy storage forms (e.g., dried silica gel or zeolite for dehumidifying, or ice for refrigerating), given that some storage techniques (e.g., drying silica gel) are more efficient than others (e.g., storing in chemical batteries, then converting to electricity, then using that electricity to drive a typical appliance).   In addition to an algorithm that decides when to store the harvested energy and in what forms, we will also investigate the sharing of energy between households in a microgrid.  Toward this, we will simulate different household energy consumption profiles, and create an algorithm for sharing among the various households.

reidReid Simmons
Research Professor Robotics Institute Carnegie Mellon

heatherHeather Knight
Graduate Student
Robotics Institute Carnegie Mellon


Conveying Robot Attitude via Motion
The same robot can perform a task in varied ways that communicate different internal states. If you eat a bowl of cereal on Thursday morning shortly before an important meeting, your motion will convey a very different attitude toward time versus eating that same bowl of cereal on a sunny Saturday with nothing on the calendar.  In the same way, a robot can convey information to bystanders about its current task state that may influence whether people engage the robot or let it pass by uninterrupted. Participants will learn about motion features that can convey varied internal states, craft several motion variations on a real robot doing a task that matters to people, then observationally evaluate diverging human response in a public space on campus.

aartiAarti Singh
Assistant Professor
Machine Learning Department
Carnegie Mellon


timothyTimothy Verstynen
Assistant Professor
Dept of Psychology and Center for Neural Basis of Cognition
Carnegie Mellon

Computational Neuroscience
The ability to predict the brain response is the key to understanding it. This project will use machine learning approaches to map the wiring diagram of the human brain (i.e., the connectome) and understand the predictive utility of these connections i.e. how well can a connectivity pattern predict future brain activity. This involves leveraging recent advances in machine learning and human neuroimaging to understand how brain networks work.

raviRavi Starzl
Systems Scientist
Language Technology Institute
Carnegie Mellon

The Future of Software Systems in Healthcare and Medicine
Participants will discuss and analyze some recent software trends in healthcare and medicine. Participants will brainstorm and iterate through a research challenge related to these fields.

Portia Singh (formally Taylor), Ph.D.
Research Staff
Acute Care Solutions,
Philips Research North America

Developing for Golden Years:  Technology for the Elderly
A worldwide “graying” of the population will lead to increased financial burden on both personal/patient and public resources.  In America alone, 20% of the population will be over the age of 65 by the year 2030.  People are living longer lives despite the physical and mental changes that occur during aging.   This project will focus on technology for the elderly that can assist in assuring a high quality of life during the later years.  We will survey current technologies, explore barriers to aging and look at the financial costs of healthcare that is associated with aging.  Students will take what they learn and propose an application to address a problem concerning elders.

The workshop will provide opportunities for all participants to work in teams on exploratory research problems. Each team will be led by a researcher from industry or academia who will introduce the research problem and guide the team through the process.

There are several sessions devoted to the research workshops.
The final research session will include presentations/solutions from each team.

For questions about the workshop, see contact page