GANs consist of three terms Generative Adversarial Network. Let's understand these three terms first. Generative : A Generative Model takes input training sample from some distribution and learns to represents that distribution. Adversarial : It basically means Conflicting or Opposing. Networks : These are basically neural networks. So,Generative Adversarial Networks are deep neural network architecture comprising of two neural networks compete with each other to make a generative model. A GAN consist of two class models : Discriminative Model :- It is the one that discriminate between two different classes of data.It tries to identify real data from fakes created by the generator Generative Model :- The Generator turns noise into an imitation of the data to try to trick the discriminator Mathematically, A Generative Model 'G' to be trained on training data 'X' sampled from some true distribution 'D' is the one which, given some standard random distrib
Image
An Image consists of a set of pixels, which are the buildings blocks for any image. Every Pixels defines the color or the intensity of light.
Suppose an image has a resolution of 1000 x 750,which mean that it is 1000 pixels wide and 750 pixels tall. So the total number of pixels in our image will be 1000 * 750 = 7,50,000 pixels.
An Image can be of two type :-
- Grayscale
- Color
A Grayscale image can have a pixel value between 0 and 255, here 0 means the pixel is 'Black' and 255 means the pixel is 'White'. All the values in between represents various shades of gray. The matrix obtained from a Grayscale Image is 2-Dimensional ie it has width and height.
A Color image is represented in RGB color space.The matrix obtained from a color image is a 3D matrix with parameter of Width, Height and Depth. Pixels in the RGB color space are no longer a scalar value like in a grayscale/single channel image – instead, the pixels are represented by a list of three values: one value for the Red component, one for Green, and another for Blue.
Before feeding the images to a neural network the image processing is required for scaling the images.The size/aspect ratio of the set of images should be same.The Common choices for width and height image sizes inputted to Convolutional Neural Networks include 32×32, 64×64, 224×224, 227×227, 256×256, and 299×299.
Loading an Image using OpenCV Library
import cv2
image = cv2.imread("example.png")
print(image.shape)
cv2.imshow("Image", image)
cv2.waitKey(0)
*pip install cv2 if module not found
Comments
Post a Comment