Time: Tuesday, May 30th @ 8:00am
Location: Dover BC
Title: "Deep Learning for Internet of Video Things – Hype or Hope?"
Abstract:
Internet of video things can collect a massive amount of video data. If the contents are properly analyzed, a great amount of information and commercial value can be extracted. To minimize human involvement, many people have high hopes for using deep learning. This is because deep learning has demonstrated the possibility of solving many difficult problems in recent years. In this panel, experts working on different areas will share their views about the collected vision of deep learning for Internet of Video Things.
Some panelists will explain why the optimism might be overly simplified and the enthusiasm might be exaggerated. For example, doubt abounds about the limitations of deep learning, in terms of the amount of computation on the edge devices, the amount of bandwidth to the cloud, and the accuracy in real world applications. Meanwhile, some panelists will provide evidence showing the significant progress in addressing the challenges, e.g., innovative devices/circuits for low-power machine intelligence, new wireless communication protocols/circuits for higher bandwidth and shorter latency, new video compression/summarization technologies to reduce communication requirements, etc.
Is deep learning for the Internet of Video Things hype or hope? You will judge.
Moderators:
1. Dr. Yen-Kuang Chen (Intel Corporation)
2. Prof. Eduard Alarcon (UPC)
Panelists:
1. Prof. Magdy Bayoumi ( the University of Louisiana at Lafayette)
2. Prof. Shao-Yi Chien (National Taiwan University)
3. Dr. Shipeng Li (Cogobuy/IngDan)
4. Prof. Yung-Hsiang Lu (Purdue University)
5. Prof. Tokunbo Ogunfunmi (Santa Clara University)
