DEV Community

Cover image for I need to learn CNN
Danh Hoang Hieu Nghi
Danh Hoang Hieu Nghi

Posted on

I need to learn CNN

#deeplearning #tensorflow #python

I need to learn CNN for my School Project presentation.
The blogs will split into 3 parts:

  • High-level overview (non-coder/non-tech background)
  • Detail how CNN work? (with visualization website)
  • Python coding example with Pytorch / Tensorflow

1. High-level overview

One of the most impressive forms of ANN architecture is
that of the Convolutional Neural Network (CNN). CNNs are primarily used to solve difficult image-driven pattern recognition tasks and with their precise yet simple architecture, offers a simplified method of getting started with ANNs.
Abstraction- An Introduction to Convolutional Neural Networks paper - For the rest of this blogs, i will try to simplify this paper and summary paragraphs and connect ideas.

Simplify this abstraction: CNN is a "better type" of ANN for image pattern recognition

Keywords: Pattern recognition, artificial neural networks, machine learning, image analysis.

Introduction

Artificial Neural Networks (ANNs) mimic human brain - nervous systems operate. ANNs comprised of interconnected computational nodes (neurons) - Learn from the input in order to optimise its final output

A simple three layered feedforward neural network (FNN), comprised<br> of a input layer, a hidden layer and an output layer. This structure is the basis<br> of a number of common ANN architectures

Load the input (multidimensional vector) fed the data to -> hidden layers. Then hidden layers make decisions from previous layer and weigh up how a stochastic (randomly) change within itself detriments (decrease) or improve the final output This called learning.

If many of hidden stacked upon each-order -> deep learning

Top comments (0)