Let’s construct an Epik-style yearbook picture changer utilizing Node.js and AI
Nostalgia is a strong power in social media, and the most recent retro development goes again to highschool with AI-generated yearbook-style pictures. Apps like Epik have gone viral on TikTok and Instagram by letting customers rework their selfies into stylized portraits that seem like they got here straight out of a ’90s highschool yearbook.
In this text, we’ll construct our personal fundamental yearbook picture changer in Node.js utilizing Catacolabs’ Yearbook Pics AI mannequin, which is hosted on Replicate. Our implementation will allow us to flip a headshot right into a ’90s-style yearbook picture. Retro!
We’ll learn to:
- Set up a Node.js venture with the Replicate consumer
- Make requests to generate yearbook-style pictures from a equipped enter URL
- Tweak the mannequin parameters for personalization
- Render the end in a really fundamental frontend
And I’ll present the complete code on the finish. Let’s dive in and trip the retro picture development.
For this information, we’ll be utilizing the Yearbook Pics mannequin to rework enter pictures into their yearbook equivalents. The Yearbook Pics mannequin is a inventive image-to-image AI that stylizes portraits within the type of ’90s highschool senior pictures.
It takes a headshot picture as enter and outputs a remodeled picture with numerous retro filters, grain, colours, and different tweaks that give it that traditional yearbook vibe.
Under the hooBuilding Our Epik-Style Yearbook Photo Appd, it makes use of a Generative Adversarial Network (GAN) structure optimized for performing this particular type switch job.
The mannequin was skilled by Catacolabs and made conveniently out there through the Replicate API.
All proper, let’s dive deeper into the technical implementation of the Epik-style app utilizing the yearbook-pics mannequin by catacolabs.
