Originally published byFierce Electronics
That is quickly becoming a mantra in the world of AI processing, where smarter devices are gathering more data than ever, and need it processed with minimal latency. It is especially important in applications where AI analytics need to be applied to data from image sensors.
With that in mind, Visionary.ai this week unveiled a real-time video denoiser to improve video image quality in the same way that noise reduction already is being applied to still images to improve their quality. The Tel Aviv, Israel, company also offers a software-based image signal processor (ISP) and True Night Vision, a technology that can enhance still images coming from LiDAR sensors. The denoiser uses an image sensor’s raw output data, which has not been compressed or degraded by any post-processing, and is designed to work alongside the company’s ISP.
Oren Debbi, co-founder and CEO of Visionary.ai, told Fierce Electronics via email,
“The Denoiser we are launching today is the next stage of our work on night vision. We now have a production-ready solution, that is power and compute efficient, and that can run on low cost silicon. We’re also learning, from our work with customers, that the denoiser technology is very relevant to IR systems which are particularly prone to noise.”
Other approaches to noise reduction of images are not fast enough to be used on live video in real time, according to Visionary.ai. The company’s approach is to do all of the image processing involved through software-based, algorithm-driven processes, in real time and at the edge, so the cost and latency involved in cloud-based AI processing is not an issue. Debbi said software-based processing helps improve the capabilities, value, and life expectancy of existing hardware, while also help save on the cost of requiring additional ISP hardware.
The company said its denoiser can be applied to extend the operating conditions for the majority of the about 7 billion image sensors that are manufactured each year.
"In very low light, when there are few photons for an image sensor to capture, noise is the limiting factor,” said Yoav Taieb, CTO of Visionary.ai, and co-founder along with Debbi. “For human vision applications, this noise adds speckles, blurs, and distortion to images, and for machine vision it reduces the accuracy of object recognition.It is not feasible to simply capture, say, ten times more photons because that would need an image sensor and lens that is ten times bigger, driving up costs.An AI based approach, that uses the raw image data and uses a sophisticated algorithm to separate the noise from the image signal is a more effective way to extend camera performance.”
Visionary.ai’s announcement comes as AI is rapidly finding its way into numerous industries and enterprises in a variety of applications. Just in the realm of image processing, AI analytics are being used to improve medical imaging, surveillance footage, live broadcast video, robotics, and more. Research firm IDC recently noted in its AI Tracker forecast that AI-related revenue from offerings that include software, hardware, and services will reach nearly $450 billion this year. IDC also said in its AI Spending Guide that more than $300 billion will be spent worldwide on software, hardware, and services for AI-centric systems in 2026.