While working on 2D to 3D converting, we received an order for changing faces of 20 people on photos. So we started exploring open source deepfake, figured out the limitations and what has to be done to achieve accuracy. This is how the first model for face swapping was born.

Our experiments with ML have been going on for years before we launched the first product. We’ve been working on sentiment analysis in texts, created a system of language dynamics analysis, and multilingual language tone analysis.We let our first face swap app out into the world. Based on advanced GAN technology, Reflect seamlessly transfers faces across images. We’ve done quite a successful PH launch, which led to a nomination for Golden Kitty Awards in the category AI & ML. Also, got a retweet from Elon Musk.

The project with photos was only a beginning, though. We didn’t want to stop there and continued refining our face-swapping model. The next step was rolling it out on video and gifs. After a lot of testing—on our faces, of course—the improved model saw the world.

We packed the gif model into our new face-swap app for gif lovers — Doublicat. The app allows you to swap faces to gifs, and instantly share them. This launch brought us mentions by media like The Verge, Mashable, Forbes, and TNW, as well as a Product of the Week award on Product Hunt.

With two successful products and an advanced face-swapping model under our belt, we decided it was time to become something bigger. So we basically made the technology we’re working on into a separate brand. This is how RefaceAI evolved.

Today, we’re actively establishing new business partnerships across game dev, marketing, advertising, filmmaking, and other industries related to content creation, while also continuing the AI/ML magic to make our technology grow.