Alex Net is the part of CNN (Convolutional neural network). Before going through the Alex Net our job is to understand what is a convolutional neural network. Convolution is the process which is mainly used for applying a filter over an image or signal to modify it. Convolutional neural networks are one of the part of neural networks, in that many hidden layers are present like convolutional layers, pooling layers, fully connected layers, and normalization layers.
Alex Net ARCHITECTURE
Alex Net uses the GPU to boost performance.
- Alex Net architecture consists of different layers like 5 convolutional layers, 3 max-pooling layers, 2 normalization layers and fully connected layers, and 1 softmax layer.
- Every convolutional layers have convolutional filters and a nonlinear activation function ReLU.
- The pooling layers will perform the max pooling task.
- Due to fully connected layers the input size is fixed.
- The input size is mentioned at most of the places as 224x224x3 but due to some padding which happens to be 227x227x3
- Alex Net consist of overall 60 million parameters.
AlexNet can be used as multi-GPU training by putting half of the model’s neurons on one GPU and the remaining half on another GPU. Not only does this mean that a bigger model can be trained, but it also cuts down on the training time.
Some of the important features of Alex Net are: –
- Overlapping Maxpooling
- ReLU Nonlinearity
- Reducing Overfitting
- Data Argumentation