Learn Before
Relation
Convolution and Pooling as an Infinitely Strong Prior
We can compare convolution to a fully connected net, but with an infinitely strong prior over its weights:
- the weights for one hidden unit must be identical to the weights of its neighbor but shifted in space.
- the weights must be zero, except for the small, spatially contiguous receptive field assigned to that hidden unit.
Pooling is similar:
- contains only local interactions and is equivariant to translation
- each unit should be invariant to small translations.
0
1
Updated 2021-07-01
.png&w=96&q=75)
Tags
Data Science
Related
Pros and Cons of CNN Architecture
Fully Connected Layer - Classification
3D Visualization of a Convolution Neural Network
Three classic networks
Convolution Filter (Kernel)
Convolutional Layer (ConvLayer)
Pooling Layer in Convolutional Deep Learning
Example of a Convolutional Neural Network Architecture
ResNets Convolutional Neural Network
Classic Convolutional Neural Network Architectures for Object Detection in Images
Inception Network (GoogLeNet)
Architecture Design
Convolution and Pooling as an Infinitely Strong Prior