This course teaches learners how to implement a Deep Convolutional Generative Adversarial Network (DCGAN) in Pytorch to generate new digits using the MNIST dataset. The course covers the idea behind GANs, the original GAN paper, the DCGAN paper, implementing the Discriminator and Generator, initializing the network, dataset, and hyperparameters, setting up the training phase, and training the network while visualizing results. The intended audience for this course includes individuals interested in deep learning, GANs, and image generation using Pytorch. The teaching method involves a tutorial-style video with a step-by-step explanation of the implementation process.
Overview
Syllabus
- Introduction
- Overview of the idea behind GANs
- Original GAN paper overview
- DCGAN paper overview
- Implementation of the Discriminator
- Implementation of the Generator
- Initialization of the network, dataset and hyperparameters
- Setting up the training phase
- Training the Network and visualizing results
Taught by
Aladdin Persson
Related Courses
-
Deep Learning with PyTorch : Generative Adversarial Network
-
Building Generative Adversarial Networks
1.0 -
Generate Synthetic Images with DCGANs in Keras
-
Build Basic Generative Adversarial Networks (GANs)
-
PyTorch for Deep Learning - Full Course / Tutorial
-
Literacy Essentials: Core Concepts Generative Adversarial Network
