Shih-Fu Chang

Chang received his bachelor’s degree in Electrical Engineering from National Taiwan University in 1985 and Ph.D. in Electrical Engineering and Computer Science from University of California at Berkeley in 1993. After Ph.D., he joined Columbia University as an Assistant Professor. He served as the Chair of Electrical Engineering from 2007 to 2010 and received joint appointment in Computer Science in 2011. He served as a co-PI and later Co-Director of Columbia University’s ADVENT Industry Consortium including more than 25 industry sponsors in the area of media technologies from 1993 to 2003. He became the Senior Vice Dean (2012-2015) and later Senior Executive Vice Dean (2015-date) of Columbia Engineering School, assuming a major role in School’s efforts in Strategic Planning, Special Research Initiatives, Faculty Development, and International Collaboration.

Chang is noted for his influential work in multimedia information retrieval, with broad applications in large-scale image/video search, mobile visual search, image authentication, and information retrieval with semi-supervised learning. His research has resulted in more than 10 technology licenses to companies and creation of three startup companies. As of August 22, 2017, his publications have been cited more than 41,000 times with an h-index of 100.

Chang’s notable awards include:

• Most Influential Scholar in the Field of Multimedia, 2016
• Honorary Doctorate, University of Amsterdam, 384th Anniversary, 2016
• Great Teacher Award, The Society of Columbia Graduates, 2013
• Technical Achievement Award, IEEE Signal Processing Society, 2012
• Technical Achievement Award, ACM Special Interest Group in Multimedia, 2011
• Fellow, AAAS (the American Association for the Advancement of Science), 2010
• IEEE Kiyo Tomiyasu Award, 2009
• Fellow, IEEE, 2004

Chang’s research includes multimedia information retrieval, computer vision, machine learning, and signal processing. The primary focus of his work is on development of intelligent methods and systems for extracting information from visual content and multimedia that are prevalent in large archives and live sources. In the early 90's, his group developed some of the earliest and best-known content-based image search systems, VisualSEEk and VideoQ, which set the foundation of this vibrant area. During last two decades, he has made significant contributions to the field of multimedia retrieval by developing large multimedia ontologies, large libraries of visual concept classifies, and automatic methods for multimedia ontology construction. These have strongly influenced design of the video search systems used in practice today. He has developed several well-known compact hashing techniques for efficient search over billion-scale image databases. His compact hashing work has enabled order of magnitude speedup and storage reduction in high-profile applications such as an online human trafficking crime fighting system (joint work with Svebor Karaman) that has been deployed in 200+ law enforcement agencies. In addition, he has developed a series of fundamental methods of graph-based semi-supervised learning that successfully address the challenge of training large-scale multimedia retrieval systems with noisy and sparse labels. These methods have been adopted in building the first commercialized brain machine interface system for rapid image retrieval. The graph-based search process, based on the random walk with restart theory, developed jointly with X. Wu and Z. Li, has also been deployed in the large app recommendation system of Huawei (connecting 1/2 billion apps to 300 million users).