报名链接：http://meetu.ps/e/BSpqZ/xKc5Q/f
Welcome to “AI Tech Day @ Palo Alto”, and this is a meet-up group initiated by “Amino Capital Fellowship Program”, for Stanford students, Silicon Valley engineers and entrepreneurs, to learn more about the technological advancements and application of artificial intelligence. This month we proudly present "Orbeus and Computer Vision Tech Day".
Dr. Huican Zhu fromAmino Capital and CEO of our portfolio company Yi Li from Orbeus (Acquired by Amazon) will give welcome remarks and share their thoughts on the future of computer vision and AI.
Mu Li, Principal Architect at Baidu, and deep learning expert at CMU, will give a talk on "MXNet: Flexible Deep Learning Framework from Distributed GPU Clusters to Embedded Systems"
Orbeus scientists, Meng Wang, Wei Xia, Yushan Chen, and Weixin Wu, will dive into "Large-scale facial and object recognition, from research to industry"
Agenda:
1:30pm - 2pm: Networking
2pm - 3:30pm: Presentations
3:30pm - 4pm: Wind down
•[Welcome Remark]
Huican Zhu, PhD (Amino Capital, early Googler starting in 2000)
Bio: Dr. Huican Zhu is Founding Partner at Amino Capital. He got his B.E degree at Sun Yat-sen University, M.S degree at National Academy of Science and PhD degree at University of California, Santa Barbara. He is the pioneer at Google image search, and the first Senior Chinese Engineer and Senior Manager at Google and has received highly praise for his excellent computer vision research capability. Huican went to Google at 2000 and took responsibility of many kinds of significant Google researches. At Google, Huican left and spent two years at Tencent during 2010-2012 to be the Chief Architect and General Manager of Infrastructure. Huican has a strong academic reputation, and his research results have over 12,000 citations. Dr. Zhu is currently a board member of MaxFun, Weee, Virtulus, Yiqixie, and Evertoon.
•[Welcome Remark]
Yi Li(Orbeus / Amazon)
Bio: Yi is passionate about leveraging technology to build creative products that people enjoy using. Yi served as CEO of Orbeus, and was featured by Huffington Post for "Women in Business". Founded in 2012, Orbeus has developed what could be the best API and application yet for letting everyday consumers benefit from advances in deep learning. Yi spearheaded the successful engagement of customers for their ReKognition API and managed the creation of a novel image recognition application called PhotoTime. Company got acquired by Amazon in 2015.
•[Presentation] MXNet: Flexible Deep Learning Framework from Distributed GPU Clusters to Embedded Systems
Mu Li (CMU, Baidu)
Abstract:This talk will describe how to develop and deploy deep learning applications efficiently and easily using MXNet. MXNet is a new deep learning framework developed by collaborators from over 10 institutes. It is designed for both flexibility and optimized performance, with easy to use interfaces in currently 7 programming languages including Python, Scala and R. We will discuss the technologies to scale out the framework to distributed clouds ranging from EC2, Azure, GCE to Spark clusters, and also memory optimizations to fit into embedded systems like mobile phones. Finally, we'll demonstrate deep learning applications in computer vision, natural language processing, and speech recognition.
Bio: Mu Li is currently a final year Ph.D. student at Carnegie Mellon University. His research interests lie in algorithms and systems for distributed machine learning and deep learning. In particular, he designs algorithms and systems scaling to petabyte datasets and running over thousands of machines. He has co-authored tens of top journal and conference papers ranging from learning theory to machine learning, from data mining to systems. He has served as principal architect at Baidu and co-founded several machine learning startups.
•[Panel] Large-scale facial and object recognition, from research to industry
Panelist:Meng Wang, Wei Xia, Weixin Wu(Orbeus / Amazon)
Moderator:Yushan Chen(Orbeus / Amazon)
Bio:
Meng Wang, Lead Scientist at Amazon
Meng co-founded Orbeus with the vision to use face and image recognition technology to quickly find photos and key sections of videos. Meng has 10+ years of experience in computer vision and image processing, including roles at Google Research, Google Search and YouTube. At Orbeus, Meng developed a high-speed, large-scale Convolutional Neural Networks (CNN) training process by building a multi-GPU system. He also developed the a high-throughput online face and concept recognition parallel computing framework that is at the core of Orbeus’ products and services. Meng studied in Boston University and conducted research at MIT and MERL, and is credited with more than 20 patents in this area.
Wei Xia, Sr Research Scientist, Amazon
Wei is now a Sr. research scientist at Amazon. His research interests focus in large-scale visual recognition and deep learning. Previously he is the Chief Scientist at Orbeus, leading the efforts of developing recognition algorithms, to support both the APIs and consumer-end products, like PhotoTime. Wei obtained his Phd in computer vision from National University of Singapore and conducted research at Panasonic and Lund University, etc.. He has published 10+ papers and served as reviewer in top-tiered conferences and journals, like PAMI/ICCV/CVPR, etc.. He has won the winner awards of PASCAL VOC classification and segmentation competitions in 2012, runner-up award in ILSVRC challenge 2013, and winner award of ISVLRC detection challenge in 2014.
Yushan Chen, Sr Product Manager, Amazon
Yushan is the product lead for PhotoTime. Prior to Orbeus, Yushan was a researchassistant in the Robotics Lab at Boston University. She is credited with inventingtheoretical and computational frameworks for automatic synthesis of distributed controland communication schemes for robotic teams from human-like high-levelspecifications. Yushan is passionate about product design and development and is atrained roboticist with the dream to improve people's daily life with smart devices, cloudcomputing and artificial intelligence. She received the “best paper” award for this workat DARS 2010. Yushan holds a Master & PhD in Computer Engineering from Boston University.
Weixin Wu ,Sr.Engineer/Tech Lead/Manager
Prior to joining Orbeus, Weixin is with Amazon’s Lab126, where he lead and developed the GPU Computer Vision Engine for both Android and iOS. He was also responsible for the Android and iOS demo prototyping, GPU architecture and driver performance. Prior to Amazon, Weixin worked at Qualcomm, where he architected and developed the world's 1st commercial mobile GPGPU OpenCL profiler and Kernel Analyzer, which is now used by Qualcomm, Samsung, Amazon, SONY and the general community of Android developers. He also served as GPGPU expert in Qualcomm’s Multimedia R&D facility, where he specialized in FastCV GPU acceleration, and achieved 5X+ faster performance against a hand optimized Neon SIMD assembly. Prior to Qualcomm, Weixin worked on AMD’s GPU Performance Application Engineering team, where he served as the judge and designed various problem sets for AMD’s APU and discrete GPU architecture, which are used for the TopCoder 1st OpenCL Competition. Weixin holds a Bachelor’s degree in Electrical Engineering from the Harbin Institute of Technology and a Master’s degree in Electrical Engineering from Clemson University.