Topic 06: Modern Convolutional Neural Networks and Adversarial Examples
This chapter introduces recent popular convolutional neural network architecture.
-
Chapter 06.01: 1D/ 2D/ 3D Convolutions
We describe convolution operation for different types of data.
-
Chapter 06.02: Important Types of Convolutions
Dilated Convolutions and Transposed Convolutions are two important types of CNNs that we introduce them in this section and describe the advantage and problem which they addressed.
-
Chapter 06.03: Separable Convolutions and Flattening
In this section, we explain separable CNNs and Flattening.
-
Chapter 06.04: Modern Convolutional Architecture - Part1
In this section, we explain the recent popular CNN-based architecture.
-
Chapter 06.05: Modern Convolutional Architecture - Part2
In this section, we explain the recent popular CNN-based architecture.
-
Chapter 06.06: Adversarial Robustness and Examples
Adversarial machine learning studies techniques which attempt to fool machine learning models through malicious input, we explain adversarial examples and adversarial training in this section.