Thank you very much for tuning in again. My name is Andreas Meyer and I want to welcome

you to our second part of the deep learning lecture and in particular the introduction.

So in this second part of the introduction I want you to show you some research that

you are doing here at the pattern recognition lab here at FAU Germany. So one first example

that I want to highlight is a cooperation with Audi and we are working here with assisted

automatic driving. We are working on smart sensors in the car. You can see that the Audi

A8 today is essentially a huge sensor system that has cameras and different sensors attached.

Data is processed in real time in order to figure out things about the environment. It

has functionalities like parking assistance. There are also functionalities to support

you during driving and traffic jams and this is all done using sensor systems. So of course

there is a lot of detection and segmentation tasks involved. What you can see here is an

example where we show you some output of what is being recorded by the car. This is a frontal

view where you can actually see that you are looking at the surroundings and you have to

detect cars. You also have to detect, and this is shown here in green, the free space

where you can actually drive. So this also has to be detected and very commonly this

is done with methods of deep learning. Today there is also a huge challenge because we

need to test the algorithms. This is done quite frequently using simulated environments.

So you use a test car on a test track and then a lot of data is produced in order to

make this somehow reliable. But in the very end this has to run on the road which means

that you have to consider different periods of the year and different daylight conditions.

This makes it all extremely hard. What you have seen in research is that many of the

detection results are working with nice day scenes and when the sun is shining. Everything

is nice so the algorithms also work very well. But the true challenge is to actually go towards

rainy weather conditions, night, winter, snow. And you still want to be able to detect and

of course not just cars but also traffic signs, landmarks and you analyze the scenes around

you such that you have a reliable prediction towards your autonomous driver system. We

also look into smart devices and of course there are very interesting topics here in

for example renewable energy and power. One problem that we typically face under production

is when there is not enough wind blowing or when there is not enough sun shining. Then

you have to fire up, backup power plants and of course what you don't want to do is over

production because that would produce energy that cannot be stored very well. So the storing

of energy is not very efficient right now. There are some ideas to tackle this like real

time prices but what you need are smart devices that can recognize and predict how much power

is going to be produced in the near future. So let's look at an example of smart devices.

Let's say you have a fridge or a washing machine and you can program them in a way that they

are flexible at the point in time when they consume the energy. You can start washing

also maybe overnight or in one hour when the price is lower and if you had smart devices

that could predict the current prices then you would be able to essentially balance the

nodes of the energy system and at the same time remove the peak conditions. So let's

say there is a lot of wind blowing then it would make sense that a lot of people cool

down the refrigerator or wash their dishes or their clothes. So this can be done of course

with recurrent neural networks and then you can predict how much power will be produced

and use it on the client side to level the power production. The example shown here is

a solar power plant. They were interested in predicting the short term power production

in order to inform other devices or to fire up backup power plants. The crucial time frame

here is approximately 10 minutes. So the idea here was to monitor the sky, to detect the

clouds, to estimate the cloud motion and then directly predict from the current sky image

how much power will be produced within the next 10 minutes. So let's say there is a cloud

that is likely covering the sun really soon then you want to fire up generators, inform

devices not to consume so much energy. Now clouds are really difficult to describe in