What?

Why & When?

  • Very effective at computer vision (e.g. recognizing cats, dogs, planes, and even hot dogs)
  • Why not feedforward neural networks (FNN)?
    • FNN do not take into account the spatial structure of the pixels
    • FNN learns global relationships in the features instead of local patterns.

How?

Keras

PyTorch

TensorFlow

Refrences

  1. Chris Albon’s note – CNN using Keras.