CNN


Notes and Ideas:

Convolutional base:

  • Used to **Extract the features ** from an image. it is formed primarily of layers performing the convolution operation.

Dense Head:

  • The head is used to determine the class of image. it is formed primarily of dense layer, but might include other layers like dropout.

visual features:

  • A feature could be a line, a color, a shape, a pattern — or some complicated combination

training the Calssifier

The goal of the network during training is to learn:

  1. which features to extract from an image (base)
  2. which class goes with what features

Transfer learning

reusing a pretrained model


Types:

U-Net architecture

TAGS