DOMAIN INSIGHT
Semantic Segmentation
How does the app know what is a person and what is background? It uses a technique called Semantic Segmentation. The CoreML model analyzes the image pixel by pixel, classifying them into different categories, allowing for precise separation of the subject from its surroundings.
CORE FEATURES
DESIGNED FOR
QUESTIONS
How does the background removal work?
One-tap AI cutout removes backgrounds using local CoreML models that detect the subject in any image. The technique is semantic segmentation: the model classifies the image pixel by pixel to separate the subject from its surroundings.
Can I combine several cutouts in one image?
Yes. Multi-image composition lets you arrange multiple cutouts on one canvas and move, scale, and rotate each of them, and several objects can be cut out from a single original image.
What can I export?
You can export cutouts as vector SVGs, replace the background with transparency, a solid color, or any image from your device, and apply more than ten styles to individual cutouts or the final composite.
Do I need a powerful Mac to run it?
No high-end hardware is required — an ordinary Mac, iPhone, or iPad can run it, and everything is processed locally with no ads, tracking, or cloud.
