Segmentation with XnorLEARN MORE
Create a new experience that literally transports you to another place
With Xnor's new image segmentation capabilities, you can take your users on a journey to new worlds or romantic getaways, or place them into their favorite games. And, with real-time image segmentation, you can dynamically isolate people in live video to blur out backgrounds and get rid of distractions.
Xnor’s groundbreaking AI performance makes this a reality for devices as small as mobile handsets and Raspberry Pi like devices. In fact, Xnor's image segmentation algorithms can be 10X faster and 10X less power consumptive, or have a 10X smaller memory footprint with no compromise in accuracy.
Overcoming the challenges of segmentation:
With real-time image segmentation you can dynamically isolate people in live video and superimpose them anywhere in 2D, VR, or augmented reality.
With image segmentation, you can blur or change the background for privacy, place people in a virtual conference room or highlight the person speaking when there is a room full of people.
Contact Xnor to learn more and see how we can help you make smarter edge devices & apps.
The Leading Edge
Xnor’s industry-leading efficient core neural network model is the fastest and and most accurate in the industry. Efficient enough to run on mobile handsets and process live camera video, Xnor’s segmentation solution runs up to 9x faster than standard solutions.
A Ubiquitous Solution
Capable of running solely on the CPU of devices or servers, it can also take advantage of GPUs and AI accelerators if available for video conferencing, streaming, and post-processing tasks.
Powerful Training Models
Similar to object detection, which classifies objects in an image and identifies its location - which you often seen output as labeled bounding boxes, segmentation takes this further by partitioning images and video frames into distinct regions containing pixels of an instance of an object. These attributes are derived by training models with images to identify different types of objects like people, vehicles, and animals.
Unlike object detection, segmentation output is much more precise - creating a binary mask of an image represented by a black and white image showing where the segmentation algorithm finds a match.