Two Information Sciences Institute (ISI) summer interns, Chenyou Fan and Katayoun Neshatpour, have been named ISI’s Director’s Interns, 2017, in recognition of their exceptional research and dedication to creating innovative, impactful and potentially game-changing technologies.
Fan and Neshatpour were two of 89 applicants for this year’s program, which ran from May until August at ISI’s headquarters in Marina del Rey, California, and the Institute’s research facility in Arlington, Virginia.
Now in its second year, the annual program welcomes enthusiastic, qualified graduate students and highly talented undergraduates to work with ISI’s senior research leaders and their teams on groundbreaking computer science research.
ISI summer interns are eligible for consideration as Director’s Interns based on their research interests, credentials, achievements and recommendations, with selections made by a committee of senior ISI researchers.
Find out more about this year’s winners:
A PhD student from Indiana University with a research interest in computer vision and machines learning, Fan worked with Wael Abd-Almageed and his team at ISI’s Arlington, Virginia, research facility.
During his internship, Fan helped to develop a deep- learning-based prototype system to improve the resiliency of face presentation attack detection (PAD) technologies.
Face recognition systems, which are widely used for entrance and access control, are vulnerable to print and video- replay spoofing attacks, otherwise known as presentation attacks. As the first shield against unauthorized access to an entire system or network, PAD is vitally important for face- recognition security applications.
“I worked on developing a two-stream deep network, which combines motion and texture features to detect spoofing attacks,” says Fan. “This method is achieving positive results on various dataset protocols, flagging 95 to 99 per cent of face spoofing attacks under different scenarios.”
Fan’s work has been published in top vision and artificial intelligence conferences, including the Conference on Computer Vision and Pattern Recognition (CVPR), the Association for the Advancement of Artificial Intelligence (AAAI) and the European Conference on Computer Vision.
A PhD student from George Mason University with a research interest in FPGA hardware design and machine learning algorithms, Neshatpour worked with Josh Monson and his team at ISI’s Arlington, Virginia, facility.
Neshatpour worked on developing techniques to apply machine learning methods to field-programmable gate array (FPGA) computer-aided design (CAD) tool flow.
Her approach centered on leveraging image classification techniques to perform circuit classification; the results would determine the optimal CAD tool configurations settings with up to 37 variables.
“We wanted to explore if it is possible to predict whether tuning a CAD tool parameter will optimize the design,” says Neshatpour. “This work could potentially help developers of digital systems reduce the amount of time they have to spend optimizing their design.”
Several of Neshatpour’s publications have appeared in high-quality conferences, including IEEE Big Data, the Great Lake Symposium on VLSI (GLSVLSI) and the ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID). She also ranked 25th out of more than 23,000 students in the Master of Electrical Engineering nationwide university entrance exam.
Interested in taking part in an ISI internship?
The application process for next summer will open in January, 2018. You can find out more, or inquire about openings for term-time internships, here.
Published on September 14th, 2017
Last updated on May 20th, 2021