
GAI NN
Blended Learning Service
Created on October 21, 2024
Over 30 million people build interactive content in Genially.
Check out what others have designed:
10 SIGNS A CHILD IS BEING BULLIED
Horizontal infographics
BEYONCÉ
Horizontal infographics
ALEX MORGAN
Horizontal infographics
ZODIAC SUN SIGNS AND WHAT THEY MEAN
Horizontal infographics
GOOGLE - SEARCH TIPS
Horizontal infographics
OSCAR WILDE
Horizontal infographics
NORMANDY 1944
Horizontal infographics
Transcript
3
4
5
2
Output Layer
7
8
6
Fully Connected Layers
Pooling Layers
Activation Functions
Convolution Layers
Input Layer
Click on the orange hotspots for more information
The structure of neural networks in facial recognition
1
Artificial neurons
Summary
In simple terms, at the lower levels of a neural network there are simple patterns. The aggregation of layers identifies more complicated and sophisticated patterns that reflect real world things, such as facial recognition. Pattern recognition, however, is dependent on training.Return to the main course to learn more about how this training works.
Summary
These functions help the network understand complex patterns by adding flexibility to the learning process. Think of it as adjusting the focus of a magnifying glass, or zooming in, to see details more clearly.
Activation Functions
The connection between neurons is assigned a weight, which determines the importance of a particular input. The more strongly connected one neuron is to another, the larger its input influence on the neuron's output. The core of the network's learning process involves adjusting these weights to minimise the difference between the network's predictions and the actual results.
Artificial neurons
Neural networks consist of layers of artificial neurons known as nodes or units, which process information. To do this, each neuron performs a simple mathematical operation.Neurons in the input layer receive raw data. In other layers, they receive input from other neurons. Neurons process the input they receive and pass an output to the next layer.
This is where the raw face image is fed into the network. Each neuron in this layer represents a feature of the input. Imagine this layer as the eyes that see the image's pixels.
Input Layer
This layer makes a decision about whose face it is. It compares the processed image to known faces and gives a probability score for each possible match to identify the person.
Output Layer
These layers act like a magnifying glass that scans the image for basic edges and textures. They look for patterns such as the shape of an eye or the curve of a mouth.
Convolution Layers
After identifying and summarising features, these layers combine all the information to form a complete picture of a faceIt's like piecing together a jigsaw puzzle where each piece is a different facial feature.
Fully Connected Layers
These layers simplify the image by picking out the most important features, similar to summarising a detailed picture into its main elements. This makes makes the processing faster and less detailed by focusing on the essential parts
Pooling Layers