6.S191: Introduction to Deep Learning

Learning and Perception

Jan Kautz; Vice President of Learning and Perception Research, NVIDIA

Speaker Bio:
Jan leads the Learning & Perception Research team at NVIDIA, working predominantly on computer vision and machine learning problems — from low-level vision (denoising, super-resolution, computational photography), geometric vision (structure from motion, SLAM, optical flow) to high-level vision (detection, recognition, classification), as well as fundamental machine learning algorithms. Before joining NVIDIA in 2013, Jan was a tenured faculty member at University College London. He holds a BSc in Computer Science from University of Erlangen-Nürnberg (1999), an MMath from the University of Waterloo (1999), received his PhD from the Max-Planck-Institut für Informatik (2003), and worked as a post-doc at the Massachusetts Institute of Technology (2003-2006). Jan is particularly interested in computational photography and imaging, computer vision, and computational displays, on which he has published many articles at various conferences including ACM SIGGRAPH, Eurographics, CVPR, ECCV, CHI, and many more.

Data Visualization for Machine Learning

Fernanda Viegas; Co-Director Google PAIR (People+AI Research)

Talk Title:
Data Visualization for Machine Learning

Talk Abstract:
Machine learning is playing an increasingly influential role in the world, due to dramatic technical leaps in recent years. But these new developments bring their own questions. What is the best way to train models and to debug them? How can we understand what is going on under the hood of deep neural networks? It turns out that data visualization can play a central role in answering these questions. I'll discuss recent work that shows how interactive exploration can help people use, interpret, and learn about machine intelligence.

Speaker Bio:
I'm a senior researcher at Google, where I co-lead the PAIR (People + AI Research) initiative and the Big Picture team. My work in machine learning with Martin Wattenberg, my long-time colleague, focuses on improving human/AI interaction with a broader agenda of democratizing AI technology. I’m also interested in weaving societal expectations and values into the design and evaluation of AI systems.

Biologically Inspired Deep Learning

Dmitry Krotov; MIT-IBM Watson AI Lab

Talk Title:
A Biologically Plausible Learning Algorithm for Neural Networks

Talk Abstract:
In spite of great success of deep learning a question remains to what extent the computational properties of deep neural networks (DNNs) are similar to those of the human brain. The particularly non-biological aspect of deep learning is the supervised training process with the backpropagation algorithm, which requires massive amounts of labeled data, and a non-local learning rule for changing the synapse strengths. In this talk I will describes a learning algorithm that does not suffer from these two problems. It learns the weights of the lower layer of neural networks in a completely unsupervised fashion. The entire algorithm utilizes local learning rules which have conceptual biological plausibility. The performance of this algorithm is comparable to the performance of standard feedforward networks trained end-to-end with a backpropagation algorithm on simple tasks.

Speaker Bio:
I am member of the research staff at the MIT-IBM Watson AI Lab and IBM Research in Cambridge, MA. Prior to this, I was a member of the research staff at the Institute for Advanced Study in Princeton, NJ. Broadly defined, my research focuses on the computational properties of neural networks. Particularly, I am interested in implementing ideas coming from neurobiology in modern AI systems. In the past, I worked on the relationship between models of associative memory (also known as Hopfield nets) and neural networks used in deep learning. I received a PhD in Physics from Princeton University in 2014.

Computer Vision Meets Social Networks

Lin Ma; Principal Researcher - Tencent AI Lab

Lin Ma is a Principal Researcher at Tencent AI Lab based in Shenzhen, China. His current research interests lie in the areas of computer vision and deep learning, such as image/video processing, analysis, and understanding. Previously, he was a Researcher with Huawei Noah's Ark Lab, Hong Kong. Lin Ma received his Ph.D. degree in Department of Electronic Engineering at the Chinese University of Hong Kong (CUHK) in 2013. He received the B. E., and M. E. degrees from Harbin Institute of Technology, Harbin, China, in 2006 and 2008, respectively, both in computer science. He was a finalist to HKIS young scientist award in engineering science in 2012. He was awarded the Microsoft Research Asia fellowship in 2011.