Back To Course Home

Generative AI with Python and TensorFlow 2

0% completed

Introduction to the Course

Getting Started

An Introduction to Generative AI

Introduction: Generative AI Applications of AI Rules of Probability Why Use Generative Models?Style Transfer and Image Transformation Quiz: Generative AI Summary: Generative AI

Building Blocks of Deep Neural Networks

Introduction: Building Blocks of Deep Neural Networks Perceptrons: A Brain in a Function Multilayer Perceptrons and Backpropagation Varieties of Networks: Convolution and Recursive Networks for Sequence Data Building a Better Optimiser Quiz: Building Blocks of Deep Neural Networks Summary: Building Blocks of Deep Neural Networks

Teaching Networks to Generate Digits

Introduction: Teaching Networks to Generate Digits The MNIST Database Restricted Boltzmann Machines Stacking Restricted Boltzmann Machines to Generate Images Creating an RBM using the TensorFlow Keras layers API Creating a DBN with the Keras Model API Quiz: Teaching Networks to Generate Digits Summary: Teaching Networks to Generate Digits

Painting Pictures with Neural Networks Using VAEs

Introduction: Painting Pictures with Neural Networks Using VAEs Creating Separable Encodings of Images The Variational Objective Importing CIFAR Creating the Network from TensorFlow 2 Quiz: Painting Pictures with Neural Networks Using VAEs Summary: Painting Pictures with Neural Networks Using VAEs

Image Generation with GANs

Introduction: Image Generation with GANs Generative Adversarial Networks (GANs)Training GANs Vanilla GAN Improved GANs—Deep Convolutional GAN Improved GANs—Conditional GAN Improved GANs—Wasserstein GAN Progressive GAN Challenges in GAN Training Quiz: Image Generation with GANs Summary: Image Generation with GANs

Style Transfer with GANs

Introduction: Syle Transfer with GANs Paired Style Transfer Using pix2pix GAN Unpaired Style Transfer Using CycleGAN Related Works Quiz: Style Transfer with GANs Summary: Syle Transfer with GANs

Course Assessment

Assessment: Introduction to Generative AI to Style Transfer

Deepfakes with GANs

Introduction: Deepfakes with GANs Deepfakes Overview Modes of Operation Key Feature Set High-Level Workflow Dataset Preparation Replacement Using Autoencoders Reenactment Using pix2pix Challenges Related to Deepfakes Quiz: Deepfakes with GANs Summary: Deepfakes with GANs

The Rise of Methods for Text Generation

Introduction: Methods of Text Generation Representing Text Text Generation and the Magic of LSTMs Decoding Strategies LSTM Variants and Convolutions for Text Quiz: Methods of Text Generation Summary: Methods of Text Generation

NLP 2.0: Using Transformers to Generate Text

Introduction: Using Transformers to Generate Text Transformers GPT Models Quiz: Using Transformers to Generate Text Summary: Using Transformers to Generate Text

Composing Music with Generative Models

Introduction: Composing Music with Generative Models Getting Started with Music Generation Music Generation Using LSTMs Music Generation Using GANs MuseGAN—Polyphonic Music Generation Quiz: Composing Music with Generative Models Summary: Composing Music with Generative Models

Play Video Games with Generative AI: GAIL

Introduction: Play Video Games with Generative AI

Reinforcement Learning

Inverse Reinforcement Learning

Adversarial Learning and Imitation

Running GAIL on PyBullet Gym

Training and Results

Quiz: Play Video Games with Generative AI

Summary: Play Video Games with Generative AI

Emerging Applications in Generative AI

Introduction: Applications in Generative AI Discovering New Drugs Solving Partial Differential Equations Creating Videos from Images Generating Recipes Summary: Applications in Generative AI

Course Assessment

Assessment: Deepfakes using GANs to Emerging Applications

Conclusion

Appendix

Setting Up a TensorFlow Lab Deep Neural Network Development and Tensorflow VSCode: An Integrated Development Environment Docker: A Lightweight Virtualization Solution Kubernetes: Robust Management of Multi-container Applications Kubeflow: An End-to-end Machine Learning Lab A Brief Tour of Kubeflow's Components Kubeflow Pipelines Using Kubeflow Katib to Optimize Model Hyperparameters

Training and Results

Learn how to train the actor-critic network and visualize the results.

We'll cover the following

Putting it all together

To train the actor-critic network, we apply a loss function that tries to classify expert (observation, action) pairs as $0$ s and agent (observation, action) pairs as $1$ s. When the agent learns to generate high quality (observation, action) pairs that resemble the expert, the discriminator will have increasing difficulty distinguishing between samples from the agent and expert and will assign agent samples a label of $0$ as well:

Get hands-on with 1400+ tech skills courses.