[Biography]

Bongjun Ko is a VP of Engineering at Samsung Electronics, Visual Display Division, where he leads the efforts to create and develop various software services on Samsung’s global footprint of screen platforms. Prior to joining Samsung, Bongjun was an AI Engineering Fellow in Stanford Human-Centered AI Institute, worked as a Research Scientist in IBM T. J. Watson Research Center, and co-founded NeoMTel that pioneered image- and video-based services in the text-based mobile era. He received Ph.D. degree in Electrical Engineering from Columbia University, and M.S. and B.S. degrees from Seoul National University.

[Abstract]

Recent advances of artificial intelligence and deep learning have been undoubtedly driven by a large amount of data amassed over the years, helping firms, researchers, and practitioners achieve many amazing feats, most notably in recognition tasks often surpassing human ability in several benchmarks (hence the mantra “data is new oil in 21st country”). The yield, however, doesn’t seem equally distributed to all who aspire to repeat the success of others in their respective domains, due to the data themselves. A selected few are running away with the infrastructure and the competence they’ve built over time to collect and process the data, leaving many others behind. For some, it’s a struggle to find ways how to get them in the first place, and for some others it’s about figuring out what to do with them. It’s somewhat puzzling that there seems abundance of data all around us, but at the same time, securing data that matters is often a very difficult feat. In this seminar, I would like to share some thoughts on this issue (and leave many open questions), as an engineer who’s been building data-driven solutions, and as an individual who plays the role of both data producer and consumer.

ZOOM : https://zoom.us/j/8978217407?pwd=d1pOZmF1OWlseEdZRVBpV3VuSkl3dz09