Over a million developers have joined DZone.

Using Generative Adversarial Network for Image Generation [Video]

DZone's Guide to

Using Generative Adversarial Network for Image Generation [Video]

Generative Adversarial Network (GAN) is class of deep learning algorithm, comprising of 2 networks - a generator and discriminator, both competing against ea...

· AI Zone ·
Free Resource

Insight for I&O leaders on deploying AIOps platforms to enhance performance monitoring today. Read the Guide.

A generative adversarial network (GAN) is a class of deep learning algorithms comprised of two networks — a generator and a discriminator — both competing against each other to solve a goal. For instance, with image generation, the generator goal is to generate realistic fake images that the discriminator classifies as real. The discriminator's goal is to distinguish real images from fake ones. Initially, the generator network starts off with blank images and continuously generates better images after each iteration, until it eventually begins to generate realistic images. The discriminator network takes an input of real images and the images provided by the generator network and classifies the image as real or fake. Then, the generator starts generating realistic images that are hard for the discriminator to discriminate as fake. The same algorithm is also being applied in other domains. However, based on my experiments, a lot of optimization needs to be done for large image sizes. I had to create a custom generator/discriminator network to work with an input size of 128*128 and 256*256 image pixels, and a lot of iterations to generate realistic images. The training data I used was from Indian Bird.

Here is a snippet of my talk on GAN at the Eclipse Summit Conference, which demonstrates the experiment.


TrueSight is an AIOps platform, powered by machine learning and analytics, that elevates IT operations to address multi-cloud complexity and the speed of digital transformation.

ai ,neural networks ,image generation ,image recognition ,generative adversarial network ,deep learning ,algorithm

Published at DZone with permission of

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}