From our beautifully simple audio and video technology to our patented, groundbreaking Neat Symmetry and more, discover Neat’s unique cutting-edge use of AI to help solve genuine meeting room problems for people.

Have you ever questioned the technology behind devices like Apple’s iPhone or Tesla’s cars? It might seem like magic when Siri knows how to respond to your voice or when a Tesla helps you avoid accidents, but artificial intelligence powers them.

Since Neat’s inception in 2019, we’ve leveraged AI to make your video meetings the best they can be. Our engineers design magical experiences that render the technology practically invisible, meaning people don’t have to think about it. “Everything just works.”

We asked some of our video and audio engineering leads to share their perspective on developing technology that embraces AI to “magically” solve real customer challenges, needs and wants.

How you use AI is like a chef’s secret

Anh Duc Dao, Neat’s Video AI Engineering Lead, believes using AI is like cooking a memorable meal.

Most innovations in engineering are about crafting and the intelligent use of known technologies to achieve a new experience. Many people can use the same ingredients, but how you mix them, the proportions, and the steps can make one recipe taste mind-blowing while another taste just ‘Meh!’ However, a great chef doesn’t have to reveal their precise steps to the public. People care more about the final result of the dish, how it looks, smells and tastes.

Anh Duc Dao, Video AI Engineering Lead, Neat

With that in mind, let’s discuss Neat’s use of AI in our cameras. We pioneered technology like our patented Neat Symmetry, which applies a unique blend of accurate people detection and robust post-processing to present everyone equally up close on screen for a more engaging, perceptible and personal meeting experience. Neat Symmetry has proven so innovative in the video conferencing industry that several competitors have ridden on our coattails and are now creating replica versions to varying degrees of success.

Incorporating people detection that’s still reliable for anyone at far distances in large meeting rooms and various other environments, as Neat Symmetry does so well, is no easy feat, but that’s what well-crafted, ingenious use of AI makes possible.

Captivating camera and lighting exposure

Most workspaces have various lighting conditions, including bright backlighting, making it difficult to see faces on camera. Neat goes beyond the traditional exposure algorithms that prioritize faces by incorporating our AI face detection. This yields better accuracy and a more extended range when finding faces to include in the exposure algorithm, which, in the end, gives everyone on a video call a better meeting experience.

Amazing auto framing

Having a camera that automatically follows people around is like having a cameraperson or director framing your meeting. It enables everyone to stay more engaged and allows people to move as they please, whether getting up to present something or to stretch their legs.

Tracking people without having to depend on speech or audio depends on video AI that allows for continuous framing of people in the room. Neat devices do this by leveraging neural networks to understand the human silhouette and distinguish it from other objects in the room. Also, Neat enables you to set virtual boundaries. Hence, the camera only follows the meeting participants, allowing you to enjoy distraction-free meetings or collaborative sessions in busy open-plan spaces.

Awesome audio

People hearing you is just as important as them seeing you. So, Neat Audio Processing ensures everyone can hear anyone loud and clear through human voice optimization, reducing disruptive background noises.

What makes this magical is how our deep noise suppression can minimize distracting sounds like fans, a bag of chips, or loud keyboard clicking without meeting participants even noticing. That’s the beauty of it. Other features like dereverberation enable meeting participants to hear people talking without the sound reflections in the room. Leveraging neural networks as the core of real-time audio processing makes all this possible.

Also, with Neat’s innovative audio technology, you won’t experience any uncomfortable double-talk issues, where the device’s audio pickup unintentionally squashes or mutes someone’s voice when two or more people talk simultaneously, and everything sounds jumbled up and incomprehensible.

While most other hardware manufacturers rely on decades-old signal processing algorithms for audio processing, our audio algorithm models are learning continuously on the backend, so they can adapt to changing room acoustics and different voices and noises to constantly improve.

Øystein Birkenes, Audio Software Lead, Neat

Leveraging AI to create more human exchanges

As more intelligent communication technologies rely on voice, such as meeting transcriptions for notes and closed captioning, live translations, etc., it’s essential to have audio technology that can accurately capture as much of the conversation in the meeting as possible in the most natural way. Having machine learning at the core of everything we do and cleverly applying it enables Neat to solve problems better, making meeting experiences feel more inherent, engaging, and free-flowing. Even magical!

Any sufficiently advanced technology is indistinguishable from magic.

Arthur C. Clarke, British futurist and inventor

Find out more about Neat’s pioneering devices or see them in action in our live Virtual Demo – Neat Devices and Experiences webinar. Better yet, book a demo.