UROP Research Opportunities
The Undergraduate Research Opportunities Program (UROP) cultivates and supports research partnerships between MIT undergraduates and faculty. If you have any questions please contact email@example.com or take a look at the How to UROP at CSAIL document (pdf format).
This program is available to MIT students only.
The Systemic Risk Dashboard is a project to bring together a collection of recognized measures of global "systemic risk" -- measures of financial stress and uncertainty in the global economy that could bring on a global economic downturn. This project, under the leadership of Professors Andrew Lo and Roger Stein, is being done in affiliation with the Consortium for Systemic Risk Analytics (CSRA) and brings together graduate and undergraduate students to work on specific, well-identified analytics developed in academic settings...Posted date: March 18, 2015
One way of studying the human visual system is to develop visual illusions (also called optical illusions) that probe specific aspects of visual processing. The Checker-shadow illusion is a well-known example from our lab. With current tools in 3D graphics and animation it is possible to create increasingly beautiful and powerful illusions. I’m looking for someone with artistic flair, an interest visual perception, and skill in the use of image-related tools such as Photoshop, Illustrator, Blender, and Matlab.
Contact...Posted date: March 16, 2015
MIT App Inventor has opening for UROPs this summer. App Inventor is a Web-based development environment for creating original apps for Android smartphones and tablets. The system is intended for people without programming experience, such as high school and middle school students, college non-computer science students, and hobbyists.
Our group runs a large-scale public Web service, with 3,000,000+ users, that lets people all over the world create their own mobile applications, and lets schools all over the world include mobile computing in school...Posted date: March 16, 2015
This project combines crowdsourcing with active learning in the context of speaker verification. While state-of-the-art machine learning methods for this task typically require a large annotated corpus of labeled speech, members of the CSAIL Spoken Language Systems group are investigating the scenario in which a large corpus of speech is provided, but none of the data are annotated. Instead, the system (1) actively chooses examples to be labeled and (2) obtains them from the crowd (i.e., via Amazon Mechanical Turk) for use in subsequent speaker modeling. Previous work has shown...Posted date: March 16, 2015
Faculty Supervisor: Andrew W. Lo
Project Description: The MIT Laboratory for Financial Engineering (LFE) and the MIT Center for Finance and Policy (CFP) are jointly working on a “systemic risk dashboard” to produce and publish live financial analytics for risk management. The goal is to provide real-time assessment of the systemic risk in the market. Building upon prior research, we are working to create a framework to monitor the financial system using real-time data aggregation and analysis. The main purpose of our...Posted date: March 12, 2015
Data is vital in assessing the severity of a crisis, informing organizations on how to prepare or give aid, and informing the community about an event. Mobile phones in general, and smartphones in particular, are an ideal tool for the collection of this valuable data. However, the development effort required to create smartphone applications is usually substantial. There are technical barriers to entry, and usually long development times. Because of this, traditional mobile application development has been limited in its ability to help disaster relief. Our framework drastically...Posted date: March 12, 2015
This project involves research in language understanding to support
processing both off-line user-generated content (e.g., social forums),
and on-line spoken dialogue interaction. Members of the CSAIL Spoken
Language Systems group are developing a generic language understanding
capability based on the open-source UIMA architecture that
incorporates a variety of natural language processing techniques.
They are actively investigating the use of embedded word vectors, knowledge
graphs (e.g., semantic web) to augment understanding capability on a...Posted date: March 12, 2015
This project involves acoustic modeling research in speech processing
to support both automatic speech recognition and speaker
verification. Members of the CSAIL Spoken Language Systems group are
actively developing acoustic modeling methods based on deep neural
network formulations (e.g., recurrent neural networks, convolutional
neural networks) to incorporate into their state-of-the-art ASR and SV
engines that are used for a range of tasks including multilingual ASR,
and mobile speaker verification. Depending on their particular
skill...Posted date: March 12, 2015
This project involves the development of speech interfaces for mobile
devices. Members of the CSAIL Spoken Language Systems group are
developing speech and language processing technologies that enable
natural conversational interaction between humans and machines. This
urop project would be suitable for the student interested in
developing intuitive human computer interfaces on mobile devices that
enable speech, text, and touch as input modalities. Depending on
their particular skill-set and interests, the superurop would be...Posted date: March 12, 2015
The InfoLab group at CSAIL seeks students interested in investigating and creating natural language tools for artificial intelligence. The InfoLab works on question answering, parsing, generating, and more, using both symbolic and statistical techniques. The START question answering system (http://start.csail.mit.edu) provides advanced access to information on the Web and in other databases via natural language. Introductory projects range from integrating knowledge sources to expanding automated methods to...Posted date: March 10, 2015
The traditional approach to privacy has been to prevent access to sensitive information. However, this does not prevent information leaks by those authorized to access it or by being inferred from public data. As an alternative, accountability is a means through we can address appropriate use. Information accountability makes the use of information transparent so it is possible to determine whether a particular use is appropriate under a given set of rules and that the system enables individuals and institutions to be held accountable for misuse. We are developing tools and...Posted date: March 10, 2015
Faculty Advisor: Brian Williams
The Model-based Embedded and Robotic Systems (MERS) group is currently working with the California Institute of Technology, the Jet Propulsion Lab (JPL), and Woods Hole Oceanographic Institute (WHOI) on a project called Resilient Space Systems. The project relies upon model-based reasoning in order to come up with a course of action for the spacecraft which does not exceed a user-specified risk bound. Thus, good models are one of the core building blocks of the RSS project.
The goal of this project is to...Posted date: March 10, 2015
The Haystack group develops tools that help people do a better job managing the information they care about: personal information, collaborative information, and social information. We apply techniques from user interface design, databases, and machine learning in tools that help people visualize data, organize information scraps, annotate online course materials, share content with friends, design web sites, hold better online discussions, and build web applications without programming. A list of ongoing projects can be found at...Posted date: December 10, 2014
We have an UROP availability to help in the development of new algorithms for analysis on large-scale clinical datasets. Large clinical image collections offer new opportunities for extracting clinically relevant knowledge, but also present challenges due to lower quality of images. Their analysis has many applications in population studies and clinical practice. The project will involve learning and developing algorithms and machine learning techniques to improve image quality and automatically segment and analyze clinical...Posted date: September 30, 2014
The Programming Languages and Verification Group is working on a number of research directions within the Bedrock project. Briefly, Bedrock is about reimagining the software development process to take advantage of machine-checked mathematical proofs at every stage -- we aim for the highest levels of rigor in showing that programs really do what they ought to do. Bedrock exists within the Coq proof assistant, drawing on ideas from such fields as functional...Posted date: September 22, 2014
Research Opportunities for MIT Students in Educational Technology and Mobile Computing
The MIT Center for Mobile Learning (a joint project of CSAIL and the Media Lab) has fall 2014 RAs and UROP positions for work in mobile computing and education. The research involves MIT App Inventor, a Web-based development environment for creating original apps for Android smartphones and tablets. The system is intended for people without programming experience, including high school and middle school students, college non-computer science students, and hobbyists.
Our group runs a large-...Posted date: August 14, 2014
Human cognition is frequently explored with a variety of tests that involve some
form of writing (e.g., solving mazes, drawing objects, etc.). With the ability to
easily capture digitized pen strokes (e.g., with a digitizing ballpoint pen or a stylus
and tablet), the data available from these tests has suddenly become hundreds of
times more precise, opening up new windows into investigating cognition, both
healthy and impaired. Our multi-site study, underway for the past several years,
has been collecting data at 7 sites around the US, providing a robust...Posted date: August 08, 2014
Social software has an inherent network effect, where people have to all use the same system in order to collaborate and interact online. This makes it hard for any new and creative applications to attract enough users to become useful. Even worse, because there is no single system that everyone is willing to use, it has become impossible to deploy any online tools that have the kind of universal reach we're used to with email and the Web.
We are developing "crosscloud" protocols which allow user data to be managed separately from applications, resulting in...Posted date: April 14, 2014
We at the Distributed Robotics Lab explore new ways of answering life-logging questions from mobile cameras, as part of the iDiary project.
Our project uses the coresets framework, along with computer vision and machine learning tools, to do online data mining from android and Google glass devices.
As part of the UROP you will work on building a UI for annotating user activities, and doing activity analysis from the video of the user, and use machine learning and metric learning techniques in order to get better understanding of the user activities.
Lab webpage:...Posted date: April 03, 2014
We at the Sensing Learning and Inference lab seek students who will build the next-generation 3D scanner! Our scanner uses ideas from robotics, computer vision, and computer graphics, and will explore the boundaries between accuracy, energy efficiency, and the interface of 3D scanners with high-level vision tasks.
Lab webpage: http://groups.csail.mit.edu/vision/sli/
Relevant skill-set: C++/CUDA, Arduino. Basic understanding of computer graphics / computer vision / probability...Posted date: April 03, 2014