Greetings from Munich

Last month, David Levin, a PhD student of Prof. Gonen Singer, visited Munich, Germany, to attend the annual International Conference on Industrial Engineering and Applications (ICIEA). The conference focuses on presenting the latest research, theories, and applications in the field of industrial engineering. “I presented a complementary article to one I presented during my master’s degree, which dealt with a “real” problem – how can we reduce the number of sensors in a multi-sensor system without hindering the system’s performance. At the current conference, I presented the application of the algorithm and its efficiency, and the results were quite impressive,” he shares. “I presented to an audience of dozens of international researchers and then answered their questions. I was also given the Best Paper Presentation award, which was quite satisfying, especially given that this was my first conference.”

Levin researches neural network compression. “My mission is to try to reduce the number of components within a network (for example, neurons in classic networks or filters in convolutional networks) without compromising the model’s performance and quality,” he explains. Reducing the number of components dramatically decreases runtimes and even improves network performance, and could possibly improve the network’s resistance to adversarial attacks (intentional input disruption in order to deceive the network).”

Levin (28) has a BSc in computer science and an MSc in electrical engineering, information processing and data science track. He met Prof. Singer during his graduate studies and pursued a PhD under his mentorship. “Apart from Gonen being a gifted professional who always has the best professional advice and knows how to present groundbreaking perspectives, he is a mentor through and through: from his professional guidance, showcased in developing my research skills, through promoting independence and critical thinking – always at eye level, and always with a smile. He is, without a doubt, a role model.”