Our Technology

Flaime combines world-leading research in computer vision, computer graphics, GPU hardware acceleration and machine learning with industry expertise in film and video game content production to develop an end-to-end pipeline for accessible volumetric capture and delivery.

Consumer Capture

Alongisde more traditional depth cameras, Flaime enables capture from consumer depth and Lidar enabled devices. Such capabilities are become more common in consumer grade devices, such as the recent iPad Pro, which is Lidar enabled.

Compression

Our proprietary compression algorithms enable highly compressed depth and point cloud data to be streamed in real-time.

We use a combination of media compression, general data compression, and neural-based compression algorithms to enable blazing fast transfer speeds.

We Render Differently

One of the major limitations with current volumetric systems is that the rendering is done using existing game engines. Although these engines are extrmeley mature for 3D content delivery, they are designed for rendering virtual content – i.e. content built as meshes.

We render differently – our custom volumetric rendering engine is designed and built from the ground up for volumetric video. Unlike game rendering engines, which are designed for meshes, Flaime’s engine is built for rendering points and voxels, scaling to billions of points in real-time. It also utilises world-leading research in deep learning to enable neural enhancement directly on the GPU.

Server-side Streaming

Cloud game streaming platforms, such as Google Stadia, are beginning to show that ultra low latency 3D content streaming can become a reality. We apply similar technologies to our volumetric platform.

The upcoming roll out of 5G will enable edge-based rendering, further lowering the latency and bandwidth of real-time content delivery.

This approach will ultimately mean that ultra-low latency volumetric video can become accessible on any web enabled device. 

Supported By