Best Paper Award for Grad Student
Sept. 23, 2019 - Mohammed Alnemari, graduate student in electrical engineering and computer science, won a Best Student Paper award at the 2019 IEEE International Conference on Edge Computing (EDGE) held in Milan, Italy, over the summer. Alnemari is a second-year Ph.D. student under the advisement of Nader Bagherzadeh, professor of electrical engineering and computer science.
Alnemari’s research looks at deep neural networks (DNN), algorithms loosely modeled on the human brain to recognize patterns. DNNs interpret sensory data through machine learning, recognizing numerical patterns in real-world data – images, sound, text or time series. The networks are used in artificial intelligence applications such as computer vision, speech recognition, robotics and more. DNNS require expansive computation and high storage costs making them challenging for devices that have constrained resources for edge computing (decentralized computing). In order to mitigate this issue, many methods have been applied and published in the literature in recent years.
In his paper, Efficient Deep Neural Networks for Edge Computing, Alnemari presents a two-stage pipeline approach, called filter pruning and tensor train decomposition, to reduce the storage and computation requirements of DNNs in order to more easily deploy them on the edge. “Our work demonstrates the same accuracy or just a tiny degradation of accuracy with retraining, after applying both stages,” said Alnemari.
The IEEE International Conference on Edge Computing (EDGE) is an international forum for both researchers and industry practitioners to exchange the latest fundamental advances in the state of the art and practice of edge computing.
– Lori Brandt