In a Convolutional Neural Network (CNN), what term describes the specific region in the input space that a particular feature is looking at?
Show answer
Receptive field
As we go deeper into a CNN, the receptive field of the neurons increases. This allows deeper layers to capture more global, complex patterns compared to the local, simple features captured by earlier layers.
Also asked as:
- In a CNN, what is the term for the portion of the input image that a specific neuron or feature is sensitive to?