/common_convolutions

Examining 3x3 convolutional filters that are common in pretrained object recognition networks

Primary LanguageJupyter Notebook

common_convolutions

Examining 3x3 convolutional filters that are common in pretrained object recognition networks (by first rounding each filter weight to the nearest 0.025).
Networks used (from Pytorch): resnet18, alexnet, squeezenet, vgg16, inception

Information about the top 100 filters in filters.txt

  • Includes weights, counts, and networks where the filter was used.

Note: only input convs included as of now

Example - plus sign

common convs

Example - point

common convs

Example - horizontal line

common convs

TODO (maybe)

  1. Include description of filters (interpretation of what they do).
  2. Aggregate multiples of the same filters together.
  3. Frequency by what layers they occur in (bottom 1/3, middle 1/3, top 1/3)
  4. Include output filters