 We saw focus on a couple of very exciting computer graphics applications that sort of bridge the virtual world with the real world or bring images from the real world into the virtual world. So you guys probably all familiar with this image, the Pac-Man. It's a lot of fun to play, probably some of you still play today, but it's actually quite easy to render based on today's standards. We can render much more advanced graphics today. And this is in big part because of the big improvements in processing power that we had over the years. Now processing power doubles every year or two more or less. And also memory increases, so the amount of memory that we can use also increased significantly. So this allows us to create much more interesting graphic images. However, our interest to have awe-inspiring graphics also increased. So we actually want the images to look better also. So we really haven't really solved the problem to some level. Thankfully, we've devised several algorithms also that can improve a lot of what we do and render and combine with the hardware advances to create very, very good and powerful results. So I'll focus on just two, which is about the time that we have, how to create interesting images of people and of places. So I'll start with places. The application that I'll describe is that of taking gigantic imagery. So you have a camera basically and you want to take a gigantic picture of a place. A way of doing that is to take several standard images and then combine them into a giant picture so then you can have a very nice rendering of a particular location or a place. This is the very first image captured. It was from 1820 in France. You can see that the resolution is not that good. There is a foreground. There is a background there. It's really hard to make out. And if you were going to process that, you'd be quite difficult. Although there are very few pixels, you'd be complex to process. This is an image, a more recent image that we took in Rio, which at the time was the image that broke the record for largest digital photograph, which is comprised of thousands of pictures, about 6,000 to 7,000 pictures, which are combined together into one giant image. You can see that you can capture a lot of little detail using only this one photograph. And this is a video demonstration of it that shows how you can zoom into a bird, for example, in the middle of the lagoon in Rio. And you can go to different places and see there is a church there. There's going to show over there. So it's a very interesting depiction of how that city looked at that point in time. So if you wanted to know how Rio looked in 2011, this is it. And you can take one today and see how the city progressed. So another exciting thing that you can do once you have all this imagery for a city is to combine it with 3D geometry. This is very commonly done with satellite imagery. What we're thinking about doing is try to incorporate that on single viewpoint images as well. So you can locally navigate regions of these images. And again, it's taken advantage of the fact that we have lots of data and are able to combine this in ways that can make the user navigation more exciting and simpler. The other application that I think is exciting that I'll discuss is of rendering people or faces to be more precise. So it's a little bit different. In the case of rendering people, you actually have cameras all over the person taking different photos from different angles instead of one camera taking pictures of several different areas for several different directions. So rendering people is actually quite complex and you actually want to create a 3D model of the person so that then you can render. This is one of the works that I did a while back which tried to capture the detail of the skin and getting the proper lighting of the skin is complex because of the way that light scatters on the skin. Just the rendering aspect is complex. But more exciting, there are several projects recently on how to capture humans and then use the rendering technology to render it efficiently. This was a project done by USC and the Smithsonian, which was very, very exciting of capturing President Obama. And basically what they did is they had a setup system that had several cameras from all different directions, several lights. And they configure the lights and the cameras in such a way they can get several different pictures that when combined, they can process and generate a 3D representation of a person. In this case, President Obama. And an exciting application of that is that then you can go and print that as well and you have a 3D representation of it. So like they took portraits of presidents in the past. Now we can actually have 3D representations of presidents. And you can do that for people as well. Not just for the presidents, but you can do it for regular everyday people. And over time, so as you age, you can see how your face changes. Another application I'd like to discuss based on that principle as well is that of movies. Basically, just like you can capture faces from different angles, you can also capture whole body motions in the movie Avatar. They have a whole capture system set up so that you can capture the character or the acting of the character. So then you can transfer the acting of the character to a virtual character. And that I think is very, very exciting. So we live essentially in days where we can actually capture both places and people very, very well. And we reach a level that historically we can now keep records of how people and places looked over time and compare, which was very difficult to do in the past. So that's what I have today. Thank you very much.