High Resolution Generative Adversarial Networks (GANs)

High Resolution Generative Adversarial Networks (GANs), available at $54.99, has an average rating of 4.4, with 57 lectures, based on 86 reviews, and has 1003 subscribers.

You will learn about Create a GAN capable of generating high resolution images using TensorFlow 2.0 Distribute training on a TPU or multiple GPUS Implement the R2 loss function Implement a scaled convolutional layer Implement up-sampling and down-sampling layers Implement mini-batch standard deviation to capture dataset variation Generate infinite random images from a trained generator Apply a perceptual path length filter to generated images Generate interpolations between two different generated images This course is ideal for individuals who are Machine learning developers who want to create high resolution images with GANs It is particularly useful for Machine learning developers who want to create high resolution images with GANs.

Enroll now: High Resolution Generative Adversarial Networks (GANs)

Summary

Title: High Resolution Generative Adversarial Networks (GANs)

Price: $54.99

Average Rating: 4.4

Number of Lectures: 57

Number of Published Lectures: 57

Number of Curriculum Items: 57

Number of Published Curriculum Objects: 57

Original Price: $29.99

Quality Status: approved

Status: Live

What You Will Learn

Create a GAN capable of generating high resolution images using TensorFlow 2.0

Distribute training on a TPU or multiple GPUS

Implement the R2 loss function

Implement a scaled convolutional layer

Implement up-sampling and down-sampling layers

Implement mini-batch standard deviation to capture dataset variation

Generate infinite random images from a trained generator

Apply a perceptual path length filter to generated images

Generate interpolations between two different generated images

Who Should Attend

Machine learning developers who want to create high resolution images with GANs

Target Audiences

Machine learning developers who want to create high resolution images with GANs

This course covers the fundamentals necessary for a state-of-the-art GAN. Anyone who experimented with GANs on their own knows that it’s easy to throw together a GAN that spits out MNIST digits, but it’s another level of difficulty entirely to produce photorealistic images at a resolution higher than a thumbnail.

This course comprehensively bridges the gap between MNIST digits and high-definition faces. You’ll create and train a GAN that can be used in real-world applications.

And because training high-resolution networks of any kind is computationally expensively, you’ll also learn how to distribute your training across multiple GPUs or TPUs. Then for training, we’ll leverage Google’s TPU hardware for free in Google Colab. This allows students to train generators up to 512×512 resolution with no hardware costs at all.

The material for this course was pulled from the ProGAN, StyleGAN, and StyleGAN 2 papers which have produced ground-breaking and awe-inspiring results. We’ll even use the same Flicker Faces HD dataset to replicate their results.

Finally, what GAN course would be complete without having some fun with the generator? Students will learn not only how to generate an infinite quantity of unique images, but also how to filter them to the highest-quality images by using a perceptual path length filter. You’ll even learn how to generate smooth interpolations between two generated images, which make for some really interesting visuals.

Course Curriculum

Chapter 1: Introduction

Lecture 1: Introduction

Lecture 2: Myths About GANs And the Truth About What They Really Are

Chapter 2: Architecture

Lecture 1: Generator – High Level

Lecture 2: Generator – Details

Lecture 3: Discriminator – High Level

Lecture 4: Discriminator – Details

Chapter 3: Weight Scaling

Lecture 1: Theory

Lecture 2: Conv2d

Lecture 3: Dense

Lecture 4: LeakyRelu

Chapter 4: Resampling

Lecture 1: Resampling Theory

Lecture 2: Blurring Theory

Lecture 3: Blurring Code

Lecture 4: Resampling Code

Chapter 5: Combined Resampling + Convolution

Lecture 1: Theory

Lecture 2: Downsampling Code

Lecture 3: Upsampling Code

Chapter 6: Minibatch Standard Deviation

Lecture 1: Theory

Lecture 2: Code

Chapter 7: PixelNorm and Image Conversion

Lecture 1: Pixelwise Normalization Theory

Lecture 2: Pixelwise Normalization Code

Lecture 3: Image Conversion

Chapter 8: Model Code

Lecture 1: Generator

Lecture 2: Discriminator

Chapter 9: Loss and Training Step

Lecture 1: High Level Training Overview

Lecture 2: Why the Wasserstein Loss Doesnt Scale to High Resolutions

Lecture 3: R2 Loss Theory

Lecture 4: Lazy Regularization Theory

Lecture 5: Step Function Code

Chapter 10: Using a TPU With a Distributed Strategy

Lecture 1: Theory

Lecture 2: Simple Example

Lecture 3: Tips

Lecture 4: Distributing Our Training Loop

Chapter 11: Supporting Callbacks

Lecture 1: Calling Back From the Training Loop

Lecture 2: Visualization Callback – Introduction

Lecture 3: Visualization Callback – Visualization Generator

Lecture 4: Visualization Callback – Callback Itself

Lecture 5: Checkpoint Callback – Overview

Lecture 6: Checkpoint Callback – Checkpointer

Lecture 7: Checkpoint Callback – Serialization

Lecture 8: Checkpoint Callback – Pickling the TrainingState

Chapter 12: Training

Lecture 1: Dataset

Lecture 2: Options

Lecture 3: train() Function

Lecture 4: Main Training Script

Lecture 5: Main Training Script Demo

Lecture 6: Training in Colab

Lecture 7: BFloat16

Chapter 13: Generating Images

Lecture 1: Perceptual Path Length Filter – Theory

Lecture 2: Perceptual Path Length Filter – Code

Lecture 3: Perceptual Path Length Filter – Effect On Variety

Lecture 4: Main Image Generation Script

Lecture 5: Results!

Lecture 6: Interpolations – Point to Point

Lecture 7: Interpolations – Circular

Chapter 14: Wrapping Up

Lecture 1: Creating TFRecords

Lecture 2: Conclusion

Instructors

Brad Klingensmith
Machine Learning Instructor

Rating Distribution

1 stars: 2 votes

2 stars: 3 votes

3 stars: 8 votes

4 stars: 29 votes

5 stars: 44 votes

Frequently Asked Questions

How long do I have access to the course materials?

You can view and review the lecture materials indefinitely, like an on-demand channel.

Can I take my courses with me wherever I go?

Definitely! If you have an internet connection, courses on Udemy are available on any device at any time. If you don’t have an internet connection, some instructors also let their students download course lectures. That’s up to the instructor though, so make sure you get on their good side!