Technology

metality sight

Metality Sight

Machine Learning based 2D-to-3D

3D display is an extraordinary emerging display technology. One of the typical challenges is lack of 3D content. Generating 3D content from 2D source is the most common way to tackle this challenge. The typical framework of 2D-to-3D conversion starts with depth map generation from 2D image. With the depth information, images of different viewpoints are synthesized based on the input 2D image. The multi-view images are then rendered and outputted to various stereoscopic or autostereoscopic displays. All the above processes are complex and usually composed of computation intensive algorithms. With the latest machine learning approach, the processing time and 3D quality of the 2D-to-3D conversion can be improved significantly.

Metality Sight

Foveated rendering

Human eyes have different clarity for the full field of view, clear in the center and blurry in the peripheral. When rendering an image on a VR device, it is not necessary for the entire frame to be same resolution. Foveated rendering refers to adjusting the resolution of the VR screen dynamically based on the gaze focus position, full resolution at the central and then significantly reduce the resolution in the peripheral. Foveated rendering can reduce the computing power required for VR scenes, which is critical for battery powered VR device.

spatial computing

Spatial Computing

Spatial Navigation

By leveraging Simbury UWB inno precise cm level ranging capabilities, it can provide 360-degree real-time position sensing information, allowing users to navigate complex indoor environments with privacy protected. Simbury UWB inno with high data throughput, ultra-low latency, low power and positioning is very well suited to enable real time synchronization and spatially aware manner, opening up new possibilities for communication, visualization and make a reality of intuitive spatial computing.

Spatial Computing

Spatial Audio

Simbury UWB inno can enable high-quality, low-latency wireless audio streaming. Its wide bandwidth and robust signal can support the transmission of high-fidelity audio without the need for compression or loss of quality. By accurately determining the position of speakers and listeners, Simbury UWB inno can optimize audio output and create a more realistic sound field. This can enhance wireless audio applications like wireless speakers, gaming setups, and soundbars, providing a seamless and immersive audio experience.

Foundation AI

Neuromorphic Analog Architecture

Human brains can perform extremely complex tasks with very low energy budgets. In contrast, digital computers consume a lot of energy when running cognitive-type tasks. AI and deep learning models seem to conform better to neuromorphic and analogue platforms. Neuromorphic Analog Architecture (NAA) is emulating human brain operation and neurons, so that the deep learning model can be modelled by mathematical simulation of neurons and implemented in analog hardware. This combination of neuromorphic computing and analog implementation greatly improve the power efficiency of the AI chip.

Neuromorphic Analog Architecture

Foundation AI

Binary Neural Network

Deep learning models have millions of parameters, which are encoded in bits. Binary neural network (BNN) targets to minimize the required memory and complexity of the deep learning model by quantizing the original multi-bit (e.g. 32-bit) parameters to 1-bit parameters. This low bit neural network reduces both storage and processing resource, which can greatly accelerate the inference process of the deep learning model. The performance of the BNN can be further maximized by implementing the model with the Neuromorphic Analog Architecture. This combination can provide quick inference response with low power operation.

Binary Neural Network