On the afternoonof Apr 6, 2016, Yuandong Tian, a Research Scientist in Facebook AI Research,gave a talk named " Reasoning in DeepLearning" to ShanghaiTech faculty and students. In this talk, Dr. Tian introducedthe prospect of deep learning, some of the innovations and breakthroughs.
First, Dr. Tian introduced the work theyare pursuing in deep learning. Deep learning has been proven very successful inmany applications that require advanced pattern matching, including computervision. However, it is still unclear how deep learning could be involved inother tasks such as logic reasoning. In this talk, he introduced two of ourrecent works on this direction, Visual Question and Answering and Computer Go. Heshowed that with different architecture, the mdoel could achievestate-of-the-art performance against existing approaches.
At the end of the seminar, Dr. Tian answered manyteachers’ and students’ questions.
Dr. Yuandong Tianis a Research Scientist in Facebook AI Research, working on Deep Learning andComputer Vision. Prior to that, he was a Software Engineer in GoogleSelf-driving Car team in 2013-2014. He received Ph.D in Robotics Institute,Carnegie Mellon University on 2013, Bachelor and Master degree of ComputerScience in Shanghai Jiao Tong University. He is the recipient of 2013 ICCV MarrPrize Honorable Mentions for his work on global optimal solution to nonconvexoptimization in image alignment.