“To create something from nothing is one of the greatest feelings, and I would – I don’t know, I wish it upon everybody. It’s heaven.” – Prince
What is GAN?
You might have used or heard about these very popular apps like FACE APP or PRIZMA. Ever wonder how these apps work?
These apps use Deep Learning techniques to generate artistic and photo-morphing images, and the Algorithm used to create these images is GAN (Generative Adversarial Networks).
Prisma totally transmogrifies your pictures to look like vividly colored abstract canvases from iconic artists like Van Gogh, Picasso, Levitan, and more.
Face App makes creepy, hilarious, weird, and sometimes fascinating alterations to faces.
GAN is about creating new images or composing music or doing something from scratch.
Here are a few live examples of GAN:
- This Person Doesn’t Exist – Generates photos of people that don’t exist, yet it looks realistic.
- This Resume Doesn’t Exist – Generates resumes that don’t exist, but it looks so real that it can fool anyone.
GANs have gained immense popularity, and for a good reason – It’s possible for a neural network to learn and generate data with the same statistics as the training set. For example, a GAN trained on photographs can generate new photographs that look at least superficially authentic to human observers, having many realistic characteristics.
In a 2016 seminar, Yann LeCun described GANs as “the coolest idea in machine learning in the last twenty years”.
Let’s not get into the math and tech part of GAN, as they deserve a separate blog and much deeper understanding.
You can play with GAN over here and here.
GANs for eCommerce
GANs can create Fashion Models with Custom Outfits. A new research paper from Berlin-based Fashion and tech Firm Zalando used GANs to produce high-resolution images, transferring customizable outfits and body poses from one fashion model to another. They used an architecture based on StyleGAN, a technique introduced by NVIDIA in 2018.
The GAN also reproduced fashion models with a variety of body types and outfits.
You can take a look at their paper here.
GANs for Personalized Products
Ever wondered if it is conceivable to have a product based on your favorite music?
Well, the future is approaching. It’s not quite viable yet, but neither is it impossible.
It’s possible to convert audio to image, and there’s a paper written on it: “TOWARDS AUDIO TO SCENE IMAGE SYNTHESIS USING GENERATIVE ADVERSARIAL NETWORK.” Humans can imagine a scene from a sound — the same way we want machines to do so by using conditional Generative Adversarial Networks (GANs).
You can find a few samples from here, where the Network was successful in converting audio to Images.
The next step would be to use Spotify’s API to get songs and then create a model that generates art from the audio and then apply the generated technique to the product!
Let’s say that you like “Take Care” by Drake and based on that song the model outputs the below image.
You can take this generated image and apply it to your product, which can be anything, and you have your customized product ready.
So that’s where the future is at, with products having a personalized touch of music that you like. Isn’t the very thought of it so heart-pleasing? Well, the day is not so far when you can create and own products based on the music you love.
Expect a lot more from GANs in the years to come. Thanks to the superior demand for custom products that online shoppers crave. And merchants have to do whatever it takes to create fantastic shopping experiences and more amazing personalized products, like using GAN to win over and retain customers.
While GAN is definitely the future of product customization, make sure your online store is already offering custom products right away to increase conversions and sales. Check out how brands across the globe increased their conversions and sales by providing custom products.
Get in touch with us to understand how to up the game with product customization.
Leave a Reply