What Is Convolution Neural Networks
What Is Convolution Neural Networks
➢ A Convolution Neural Networks is a type of neural networks that is used for the
image processing problem.
➢ The advancement in computer vision with deep learning has been constructed and
perfected with convolution neural networks.
➢ Some of the application of CNN are
1.Image Recognition
2.Video Recognition
3.Image Analysis
4.Image Classification
What are the layers in Convolution Neural Networks?
1.Input Layer
2.Convolution Layer
3.ReLu Layer
4.Pooling Layer
5.Fully Connected Layer
Input Layer:
➢ Input layer holds the raw input of image with width,height and depth.
Convolution Layer:
➢ This layer computes the output volume to do that it calculate the dots product
between filter and image
➢ Convolution is the first layers to extract features from an input image.
➢ Convolution is a mathematical operation that takes two input.
1.Image matrix
2.Filter
What is Pooling layer?
➢ The pooling layer is responsible for reducing spatial size of the convolved features.
➢ We add pooling layer to decrease the computational power required to process the
data.
➢ They are two types of pooling layers
1.Maximum pooling
2.Average pooling
Maximum Pooling:
➢ Maximum pooling returns the maximum values from a portion of image that
is covered by Kernal.
Average Pooling:
➢ Average pooling returns the average of all values from a portion of image
that is covered by Kernal
what is Flatten?
➢ Adding a flatten layer we are converting an array into vectors.