Fritz Image Segmentation

Create immersive camera-based apps with image segmentation

Posted Embed

Create more engaging, immersive mobile apps with Fritz Image Segmentation. Generate pixel-level masks of objects in images to create fun photography effects, improve occlusion in AR, and more.

You need to become a Contributor to join the discussion.
Jameson Toole
Jameson TooleMaker@jamesonthecrow1 · Co-founder and CEO at Fritz
Hey Hunters, We're back with another amazing ML powered feature for your apps. The more a camera can see, the more there is for users to engage with. A camera that separates background from foreground can transport someone to a different world. A camera that knows where physical objects are can immerse them in mixed reality. Apps that embrace these capabilities will attract and retain more users. Today, we’re announcing a new tool to help developers do more with the camera: Fritz Image Segmentation. Building on image labeling and object detection, image segmentation labels every pixel in an image with the object it belongs to. Developers can use these pixel masks to create better photo editing tools, video effects, improve occlusion in augmented reality, and build autonomous systems that understand and react to their surroundings. We’ve optimized deep learning models to segment three common types of scenes: outdoor streets, living rooms, and scenes containing people. As always with Fritz, everything runs on device which means it doesn’t require an internet connection and can be used in realtime on many devices. If you have a custom set of objects you’d like to mask, get in touch and we can help you with a custom model. Or try our open source training template available on GitHub: The Fritz SDK, including image segmentation, is free to use, so check it out and let us know what you think! The Fritz Team