Hello everyone, welcome back to the Computer Vision course.

In this second part of the first lecture, I'm going to continue introducing you to different

applications of Computer Vision.

In the last slide, in the last video, we were talking about 3D reconstruction techniques

in Computer Vision.

Let's continue there.

So in the last video, we saw how different angles of different angles, if you're taking

images of when you're recording a movie, if you're taking images from different angles,

you can reconstruct the whole scene and give it a circular motion, creating an impression

of a maze because how can you go around the character without actually simultaneously.

So that was one of the very good applications we saw of Computer Vision.

Another one is human-shaped capture.

So here you see President Obama.

This is an example of how a digital image or digital shape of President Obama, the previous

president of the United States was created.

So this setup is specifically for that kind of thing.

You see that there are a lot of cameras here which are not on and there are only some cameras

which are on.

So the technique requires that it records the reflectance and the reaction of your skin

color to different brightness intensities.

And therefore, a lot of images are taken with different brightness intensities to create

a realistic looking skin tone reflection of a particular person.

These cameras are not normal cameras.

These are fisheye lenses and fisheye lenses have, so these cameras have fisheye lenses.

Fisheye lenses are basically an ultra wide angle cameras.

It can capture more real world information in a particular frame.

And due to that, you see the images to be not very flat.

You will see them to be a bit spherical in shape.

It gets a larger field of view for a particular frame.

And that's the advantage of using the fisheye lenses.

Here you see that the reflectance generated different skin tones for the president.

And due to that, it was superimposed on this, superimposed on the left 3D shape generated

of the president with his color skin.

And this was the final bust of the president generated, which is clearly quite good.

This is also a very good example of how computer vision meets graphics.

So using different camera setup, different lenses, you created or you captured a lot

of information from the real world and you created a model of the real world in the image

domain.

And computer graphics helps you here by taking all this information and rendering it in a

3D format like this, for example.

Another example, another application is shape capture.

So similar to the previous example, here you see, if you don't know this character, his

character's name is Admiral Tarkin.

He has featured in Star Wars movie franchise.

However, recently he passed away.

And in order to bring back his character, the movie producers went ahead and created

something similar to what they did with digitizing President Obama's 3D shape.

They collected a lot of images of this character from the previous movie and generated a realistic

looking character in the new Star Wars movie called Rogue One.

Another example is of motion capture.