Overall Explanation

Overview

Image Recognition, in the context of computer vision, is the computer’s ability to recognize and identify an object through an image, a video or a live camera.

This allows us to solve many real world problems from governments detecting criminal amongst crowd, phone locking systems, self-driving cars and so on.

Similar to our googlenet example we can detect 1000 different objects from animals to house hold items.

AlexNet

What is GoogleNet?

Alex net is an artificial intelligence model that was developed and designed by Alex Krizhevsky in collaboration with Ilya Sutskever and Geoffrey Hinton. It adopts Convolutional Neural Network (CNN) to process the image.

When we look at a picture, each of our neurons takes a part of the image that we saw. When those neurons connect with each other, a image is processed within our brain.

CNN tries to replicate our brain, by having layers that processes simple parts of the image such as lines, and curves and combining the layers to process more complicated patters such as faces, objects and so on.

The AlexNet stacks and processes these CNN layers to create a neural network model that can identify faces, objects and others in a detailed and accurate way.

How are GoogleNet and AlexNet?

If both GoogleNet and AlexNet are using the same CNN, why are they different? Although they do use the same CNN to process the image, the number of layers, and the number of stacks and arangements for the CNN for each of the model is very different.

Similar to how camera lenses use multiple stacks of magnifying glasses with different settings to create clear picture, Both of our networks use different arangements of CNN with different additional layers and settings to deliver accurate identification of a given image.

reference: https://github.com/dusty-nv/jetson-inference/blob/master/docs/imagenet-console-2.md reference: https://towardsdatascience.com/convolutional-neural-networks-explained-9cc5188c4939