Alex achieves image classification: model principle + training + prediction
Image classification or retrieval tasks are widely used in search operations in browsers and crawler search images. This article mainly implements cat and dog classification through the Alex model, and gives a reusable open source model in the article! ! !
This article will discuss from the following contents:
1. Model introduction and environment construction
2. Dataset preparation
3. Training
4. Forecast
1. Model introduction and environment construction
Model introduction: AlexNet is the first deep convolutional neural network applied to image classification proposed by Alex Krizhevsky. The network achieved a top-5 test error of 15.3% in the 2012 ILSVRC (ImageNet Large Scale Visual Recognition Competition) image classification competition rate to win first place. Also after that year, more and deeper neural networks were proposed, such as the excellent vgg, GoogLeNet. This is already quite good for traditional machine learning classification algorithms. The network structure is as follows
AlexNetCarry LeNetforward the idea of the theory, and CNNapply the basic principle of the theory to a very deep and wide network. AlexNetThe main new technologies used are as follows: