"Most applications of deep learning use “convolutional” neural networks, in which the nodes of each layer are clustered, the clusters overlap, and each cluster feeds data to multiple nodes (orange and green) of the next layer. Credit: Jose-Luis Olivares/MIT"(1)
The deep-learning network is one of the most interesting things in machine learning. When we are talking about deep learning, we should describe what that term means. Deep learning means that the creature knows how to make something. But also the reason why that thing is done, and when it must do that thing.
In the case of deep learning, the creature knows the background for some regulations. But the thing is that machine learning allows that the AI would collect data about everything from the internet. The system can simply copy-paste the texts from the homepages and then turn that text into speech. Learning new skills means that the system connects the databases to new entireties.
So the deep learning means that the creature knows what to do when to do and where to do something. When we are thinking about deep learning from the point of view of artificial intelligence. We know that AI can make impressive things. Artificial intelligence can make things like movement series but the question is, does it know what it does?
We all can make things like a series of movements and talk about many things. And we might know nothing about the things that we are talking about or doing. In the case of machine intelligence, deep learning means that the system records the actions like macros.
For robots, the movement series are the macros that are turning to the movement series. When the robot sees the person who is injured that thing activates a series of movements. Whose mission is to save the life of that person. The thing that activates the movement series is the details like blood. Or something else that is stored in the memory of the computer.
So the cameras of the robot are sending images to the computer. And if there is a match with the accidents that thing activates the help mode. The robot might ask "are you ok"? And if the person answers "hurts" or is quiet the robot would give first aid. If the person can talk the robot would ask what happened? If the injury is caused by violence.
The robot would activate the protective mode. When the robot is searching criminals some details are activating the attack mode. Even if the robot seems like a human. It can use infrared and radar scanners. For searching things like firearms and knives. When the system detects a weapon. That thing activates the attack. The idea is that when the system detects some image.
That image activates the workspace. The operation is quite similar to operating systems in which locking can open by using an image. In that kind of system, the image activates the movement series. And the thing is that those movement series are programmed similar way. Without depending. If those movement series meant for opening the door.
Or if they meant for the use of weapons. The programming of the movement series is always similar. If the robot uses a drill machine the image of the drill machine activates the program how to use that thing. And the same thing is useful in the cases that the robot must use any other tools or weapons.
The robot can take the gun, but it can also have ultrasound or microwave systems or internal weapons like shogun in the hand bones of the artificial skeletons. The image of the gun is acting like a trigger that activates the attack mode. The thing is that the robot itself is a multipurpose machine. And the programming determines what it can do and what it cannot do.
(1)https://scitechdaily.com/deep-learning-ai-explained-neural-networks/
Comments
Post a Comment