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