Feature Extraction through CNN With Examples| Convolution, Threshold and Pooling by Dr Arshad Afridi
In this video lecture we discussed Convolutional Neural Networks also called CNN. Feature Extraction through CNN With Examples| Convolution, Threshold and Pooling by Dr Arshad Afridi
Convolutional Neural Network Learning Step Wise| Convolution, Threshold and Pooling Dr Arshad Afridi.
CNN Understanding Step-by-Step With Examples| Convolution, Threshold and Pooling by Dr Arshad Afridi
Feature Extraction via Convolution is a way to find out local features.
Convolution, threshold and pooling are the three main steps of Convolutional Neural Networks also called CNN.
Feature Extraction by Convolution in three dimension (3D).
Convolution is done through a kernel or a filter.
Pooling is the Subsampling from m by m pixels into 1 pixels
Various types of pooling are Max Pool, averaging Pooling or L^p pooling.
Subsampled feature map is extracted as a feature map.
Advantage of Pooling are Reducing the number of parameters and
Generating more robust feature maps like Shift Invariant.
Due to this Many Moreno n-zero values are matched which help against the shift invariant phenomenon.
summary of this process is:
Image - Convolution- threshold- sub sampling - gives feature map.
Feature map is also an image.
We can extract many different feature maps
If we can cascade this step many times, we can have a higher feature map.
CNN Understanding Step-by-Step With Examples, Convolution, Threshold and Pooling, Convolutional, convolutional neural network tutorial, convolutional neural network, CNN Pooling, CNN Convolutio, cnn convolutional neural network, CNN thresholding mechanism, CNN by dr arshad afridi, deep learning tutorial for beginners, deep learning with dr arshad afridi, Example of cnn, cnn implementation, cnn mnist examples, CNN layer by Layer, cnn steps down, feature extraction using cnn
#ConvolutionCNN #ThresholdingCNN
#PoolingCNN