WiMi Hologram Cloud to Introduce A 3D-CNN-based Hologram Classification Algorithm
Considering Apple has just stepped up the plate to introduce its Vision Pro mixed-reality headset to the world, it’s apparent that today’s tech consumers have plenty to be excited about. But what about the future? Whilst VR (virtual reality) headsets have allowed users to immerse themselves in new worlds, the boundless potential of AR (augmented reality) seems comparatively untapped for consumers. That is, until now. New hologram tech is on the horizon, and it sounds exciting.
As reported on PR Newswire, WiMi Hologram Cloud Inc., a global leader in Hologram Augmented Reality (AR) technology, has announced the development of an innovative new algorithm, based on deep learning. WiMi has pioneered a new 3D-CNN hologram classification algorithm – it uses a convolutional neural network and computer vision to, as the company puts it, “build classifiers for classifying targets in holograms”. Sounds complicated, right? Let’s dig in.
New hologram tech – the science inside
What this new technology promises to do is capture the shape and spatial features of a 3D target more accurately. The algorithm processes the 3D space in real-time, radically cutting down on the required processing power – the hologram image itself passes through convolutional, pooling and fully connected layers. WiMi’s algorithm passes these layers to present a more convincing vision to the user. CNN? No, it’s not just a news network, although the organisation itself has tricked passive audiences with purported hologram tech in the past.
In this context, it stands for ‘Convolutional Neural Network’, and it relies on our senses’ own 3D processing power.
3D-CNN can efficiently extract 3D features of multiple resolutions and combine them to improve classification performance. When training the model, supervised learning is performed using labelled holograms, and a back-propagation algorithm optimizes the model parameters.
– PR Newswire
How it works
Hologram tech often utilises a ‘base’ locked in place, tracked by sensors on the user’s headset. What WiMi is aiming to achieve is to provide essential technical support for object recognition, training neural network models to present fast, accurate classification of holograms. The hologram is pre-processed (turning it into a 3D image), and 3D-CNN tech learns its features.
The ultimate goal is to better handle 3D and wavefront information, presenting a hologram with high accuracy. To cut down on the native processing power of your headset and computer, it utilises GPU (the graphics processing unit found in most modern computers) power for better results.
“The 3D-CNN-based hologram classification algorithm has many applications and development prospects in several fields. Meanwhile, its technical principle can also be applied to classifying or processing other 3D images, which have good promotion value. 3D-CNN-based hologram classification technology has been widely used in autonomous driving, medical image diagnosis, intelligent security, virtual reality, etc.” – PR Newswire
So, whilst you might imagine hologram chats with your family, ‘hologram classification’ has myriad applications in the real world. Imagine, whilst driving, the tech being used to identify surrounding obstacles. It could be a breakthrough for autonomous driving. It could also serve the medical industry, analysing and diagnosing medical imagery to help doctors make smarter decisions.
We often talk about augmented reality in terms of headsets in the home, but it’s clear that WiMi’s advances could revolutionise our tech’s understanding of the world around us.
What do you think augmented reality power processing could be useful for?
Source: WiMi Hologram Cloud to Introduce A 3D-CNN-based Hologram Classification Algorithm.
Want more breaking alternate reality news? How will it shape social media? Read $6 Million AR Creator Fund Unleashed on TikTok.
