Keynotes

	Jeff Dean Google Title: Large-Scale Deep Learning For Intelligent Computer Systems Abstract: Over the past few years, we have built two generations of computer systems for training and deploying neural networks, and then applied these systems to a wide variety of problems that have traditionally been very difficult for computers. We have made significant improvements in the state-of-the-art in many of these areas, and our software systems and algorithms have been used by dozens of different groups at Google to train state-of-the-art models for speech recognition, image recognition, various visual detection tasks, language modeling, language translation, and many other tasks. In this talk, I'll highlight some of the lessons we have learned in using our first-generation distributed training system and discuss some of the design choices in our second-generation system. I'll then discuss ways in which we have applied this work to a variety of problems in Google's products, usually in close collaboration with other teams. This talk describes joint work with many people at Google.
	Yann LeCun Facebook and NYU Title: The path to AI requires us to solve the unsupervised learning question Abstract: TBA
$Description: Description: Description: Description: C:\Users\trevor\Dropbox\t3\etc\homepage\trevordarrell.jpg$	Trevor Darrell University of California, Berkeley Title: Perceptual representation learning across diverse modalities and domains Abstract: Learning of layered or "deep" representations has provided significant advances in computer vision in recent years, but has traditionally been limited to fully supervised settings with very large amounts of training data. New results show that such methods can also excel when learning in sparse/weakly labeled settings across modalities and domains. I'll review state-of-the-art models for fully convolutional pixel-dense segmentation from weakly labeled input, and will discuss new methods for adapting deep recognition models to new domains with few or no target labels for categories of interest. As time permits, I'll present recent long-term recurrent network models can learn cross-modal translation and provide open-domain video to text transcription.