Four computer science graduate students have been named Amazon Fellows, a program run by the JHU + Amazon Initiative for Interactive Artificial Intelligence (AI2AI). The fellowship supports PhD students who are pursuing innovations that advance the state of the art in interactive AI technologies.
Kelly Marchisio, Arya McCarthy, Desh Raj, and Anshul Shah were selected based on their outstanding publication record, research proposal, and mentor support.
The Fellows will receive a full stipend, 20% tuition, and student health insurance for the fall and spring semesters. Additionally, they will be nominated for a paid summer internship at Amazon, during which they will gain valuable industry insights and experiences via engagement with Amazon researchers.
Meet the Fellows
Kelly Marchisio, from New Providence, NJ, works at the Center for Language and Speech Processing and is advised by Philipp Koehn. Her research is in machine translation, particularly for languages currently underserved by language technologies. Her recent work concerns the question: If given data from two languages but no knowledge of which words are translations of one another, how can one recover the translations? Recently, she has begun examining how to efficiently make existing language technologies work for new languages.
Previously, she earned an MPhil in Advanced Computer Science from the University of Cambridge (2018), an Ed.M. in Mind, Brain, and Education from the Harvard Graduate School of Education (2013), and a BA in Psychology/Sociology from Franklin & Marshall College in Lancaster, PA (2012). Before returning to academia for graduate studies in computer science, she was a web developer at Google. Kelly enjoys classical singing and was a soprano Choral Scholar at Christ’s College, University of Cambridge. She toured internationally with the choir to Canada, Singapore, and New Zealand.
Arya McCarthy designs principled and generalizable models of language, reaching into and across families. The work spans machine translation of text, spoken language translation, and word formation processes, including an analysis of color words across thousands of languages. He is is working with David Yarowsky.
Arya previously attended Southern Methodist University, where he earned his BS in math and computer science, and his MS in computer science. While he is originally from Dallas, Texas, he has also called Edinburgh, Boston, Baltimore, the Bay Area, and New York his home. In his spare time, he bagpipes–a self-proclaimed great way to social distance–in addition to baking, biking, reading, and running.
Desh Raj works at the Center for Language and Speech Processing and is advised by Dr. Sanjeev Khudanpur.
He has worked on several things under the broad umbrella of speech processing, but his thesis is focused on building systems that can recognize free-flowing multi-party conversations. He develops models that tell you, “who spoke when and what,“ enabling applications such as real-time meeting transcription and smart medical assistants.
Desh grew up in India and graduated from Indian Institute of Technology Guwahati (IIT Guwahati) in 2017 with a bachelor’s degree in computer science. When not working, Desh splits his time doing bouldering, picking his guitar, and reading fiction.
Anshul Shah’s research interests are in the area of pose-based video understanding, self-supervised learning, and multi-modal learning. Some of his current ongoing projects include developing pose-based action recognition models to help in the early diagnosis of developmental disorders, and a project to extract key steps from procedural videos without labels, which is suited to AR applications. Shah works under the guidance of Rama Chellappa.
Anshul obtained an MS in computer science from the University of Maryland, College Park, and he received his B.Tech (Honors) and M. Tech in electrical engineering from the Indian Institute of Technology Madras. He hails from Vadodara, a city in Gujarat, India, and likes to cook and travel in his spare time.