Convolutional neural network

A convolutional neural network (CNN, or ConvNet) is a program used by computers to see things in the real world. A camera or other system takes a picture, and the computer uses the CNN to tell what the picture is or means. It is part of deep learning. Models that work with spatial data, for example images and video, use convolutional neural networks. People use convolutional neural networks with vehicles that drive and steer themselves, medical imaging, and in computer vision and deep learning.

Convolutional Neural Network Media

• 

  1D convolutional neural network feed forward example

• 

  CNN layers arranged in 3 dimensions

• 

  Input volume connected to a convolutional layer

• 

  A worked example of performing a convolution. The convolution has stride 1, zero-padding, with kernel size 3-by-3. The convolution kernel is a discrete Laplacian operator.

• 

  typical CNN architecture

• 

  Three example padding conditions. Replication condition means that the pixel outside is padded with the closest pixel inside. The reflection padding is where the pixel outside is padded with the pixel inside, reflected across the boundary of the image. The circular padding is where the pixel outside wraps around to the other side of the image.

• 

  Worked example of 2x2 maxpooling with stride 2.

• 

  Max pooling with a 2x2 filter and stride = 2

• 

  RoI pooling to size 2x2. In this example region proposal (an input parameter) has size 7x5.