[upbeat music] - Here, a student can have whatever area of AI they would like to explore.
- My area is artificial intelligence.
Particularly my dissertation topic right now, it's in the aspect of IoT or Internet Of Things, where everything is connected.
That's why we're doing this, right?
We want to live in an environment where it's secure for us, where we all feel protected and you don't have to give up your privacy for that.
- Currently I'm trying to identify approaches and methodologies to identify individuals, and this is based off of either your DNA, as well as anything within the environment that can hint as to you the individual.
I'm trying to equate evidence to one's identity.
And as the suspects align to the fingerprint evidence and to the photographic evidence, you can see how they are combined after Dempster-Shafer theory is conducted.
So these values are mass values and the one that's the closest to one, so here we subject to a one, they're the person that you would start your investigation with.
Every 39 seconds on average, a cyber attack happens, and as a cyber security professional, I have to be as innovative as the attacker and think about what is it that they are attacking?
I'm on the defense, so therefore I need to figure out, how do I solve or save that resource or that asset?
- It's really cool as a sophomore at A&T that I have the ability to work with automatic driving systems.
I feel like not many kids my age get to do that.
In America each year, there's hundreds of thousands of accidents causing death and serious injuries.
So what we wanted to do was really minimize all of this, and the only way to do this is through automatic driving system, 'cause when you eliminate the human error, the freeways will be clean and clear.
Our task in this project is what we're doing is we're analyzing all of the images coming in.
So on the self-driving car, there's three cameras, one on the side, one on the top, and one on the right.
Data is being flooded into these cameras, hundreds of pictures a second, so it's our job to make out what these pictures are saying.
So what happens is that these three pictures are flooded into what's called a convolutional neural network.
So this convolutional neural network is fed in with the desired steering command, and with this, you're always gonna get an error.
So it could be minimal.
It could be high, but we have something called backwards propagation, which goes back and corrects the error.
It'll adjust the weights to make it perfect, and so that way you don't have to worry about it failing because it'll correct it by itself and after this is done, you'll be able to almost perfectly analyze cars, pedestrians, and street signs.
- I mean, the students, they are the front and center of research.
They get excited, new knowledge, new understanding.
They can see that how they can apply how transformative they are, how translational they are.
So when I see that they're excited, you know, I get excited.
There are so many different projects in how we can use technology to transform people's lives.
[upbeat music]