Reading: Deep Roots — Improving CNN Efficiency with Hierarchical Filter Groups (Image Classification)

Similar or Higher Accuracy Than the Baseline Architectures Such as NIN, ResNet & GoogLeNet, with Much Less Computation & Smaller Model Size

For a convolution, it is unlikely that every filter (or neuron) in a deep neural network needs to depend on the output of all the filters in the previous layer. In fact, reducing filter co-dependence in deep networks has been shown to benefit generalization.

Outline

1. Convolution with Filter Groups in AlexNet

AlexNet — Convolution with Filter Groups

2. Root Module: Architecture

Root Module

3. Root Module in NIN

NIN Root Architecture
NIN Variants with Root Modules
Accuracy on CIFAR-10
The inter-layer correlation between the adjacent filter layers conv2c and conv3a
Root, Column, & Tree, for Root Module Variants in NIN

4. Root Module in ResNet

ResNet Root Architecture
Accuracy on ILSVRC
Accuracy on ILSVRC

5. Root Module in GoogLeNet

GoogLeNet Root Architecture
Accuracy on ILSVRC

It has been a long time not reading a CVPR paper about image classification.

This is the 7th story in this month!

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store