您的瀏覽器不支援JavaScript語法,網站的部份功能在JavaScript沒有啟用的狀態下無法正常使用。

Institute of Information Science, Academia Sinica

Events

Print

Press Ctrl+P to print from browser

Seminar

:::

Treat Every Joke Seriously: The Story of a Research Project

  • LecturerProf. Yung-Hsiang Lu (School of Electrical and Computer Engineering, Purdue University)
    Host: Yuan-Hao Chang
  • Time2019-12-19 (Thu.) 13:30 ~ 15:30
  • LocationAuditorium106 at IIS new Building
Abstract

The joke by two sophomores eight years ago changed the way I do research. The two students asked me whether it would be possible to see the world without paying airplane tickets. We started a new research project. Over the eight years, more than 300 students have participated in this project. I was promoted to a full professor. The research team established long-term collaboration with several companies. Academic collaborators come the College of Agriculture and College of Health and Human Sciences, in addition to College of Engineering and College of Science. Two undergraduate students received honorable mention of the North America Undergraduate Research Award by the Computer Research Association. A patent was granted. Students received four awards in business plan competitions. One group of three students started a technology company and raised more than one million US dollars. Purdue's College of Engineering is expanding the undergraduate research programs. I am assigned the Engineering Dean's Fellow for Entrepreneurship coordinating and improving the entrepreneurship programs.

This research project discovers network cameras that can provide open real-time visual data. The project considers how to use cloud computing efficiently. Efficiency is essential when processing hundreds or thousands of data streams. Cloud vendors provide many types of instances: different amounts of memory, different number of cores, etc. The geographical locations of cloud servers may affect the data rates. Efficient usage of cloud resources can be formulated to a multi-dimensional bin packing problem. The data from network cameras has different characteristics from the popular datasets for machine learning. Hence, this research team also investigates the properties and bias of datasets. Machine learning often uses data from the Internet and privacy protection has become a major concern. This research team is investigating methods to protect privacy.

The seminar will review the history of this project and discuss the major decisions over the eight years.

BIO

Dr. Yung-Hsiang Lu is a professor at the School of Electrical and Computer Engineering at Purdue University, West Lafayette, Indiana, USA. He is a Distinguished Scientist of the ACM. He received the PhD. from Stanford University and BS from the National Taiwan University. He published a book "Intermediate C Programming" (CRC Press). Computing Reviews said, "If you land on a desert island that has a Linux computer, this is the one book to have with you."