Two College of IST faculty members earn Amazon Research Awards

Professor James Wang (left) and Assistant Professor Xinyu Xing, both faculty members in the College of Information Sciences and Technology, are recipients of 2019 Amazon Research Awards. Credit: Penn StateCreative Commons

UNIVERSITY PARK, Pa. — Two Penn State researchers — both from the College of Information Sciences and Technology — are among 51 computing experts representing 39 universities worldwide to receive 2019 Amazon Research Awards.

Professor James Wang and Assistant Professor Xinyu Xing received the awards this year. For Wang, it is his second consecutive Amazon Research Award since being the first Penn State researcher to earn the recognition in 2018.

The award program provides funding at academic institutions worldwide for research in a variety of computing domains, including computer vision, machine learning algorithms and theory, and robotics.

Wang’s award will enable his research team to develop novel computational methods for their ongoing work, which focuses on how to train a computer to recognize human emotions from body movements. Funding from the 2018 award allowed the team to collect an open dataset for bodily expressed emotion understanding, which has recently been made available to the research community.

“In the past, we have taken a couple of approaches to tackle this very challenging problem and made good progress,” said Wang. “But there is much room for improving the accuracy and usability.”

He added, “We plan to expand the technological exploration in the coming year [with the new funding], leveraging the advancements in computer vision and machine learning.”

For Xing, the funding will help his team explore, design and develop solutions to combat the threat of malicious software. Additionally, through use of the Amazon cloud platform made possible with the award, the team will be able to evaluate their methods on a large scale.

“When talking about malware detection, the biggest headache in the security community is how to deal with inaccurate and sometimes incorrect data labeling,” said Xing. “This funded research will allow us to explore solutions for this acute issue. If successful, it will benefit not only the research in malware detection, but more importantly many other data-driven research because they also share the challenges we observed in malware detection.”

Last Updated June 20, 2020