Machine Learning for Imaging
Instantaneous Low Light Imaging.
In a fraction of a second, our ALLIS Neural Network and the Camera Profile merge at the edge to produce stunning instantaneous low noise imagery.
A form of Machine Learning based Artificial Intelligence, the ALLIS neural network embeds itself into virtually any image sensor.
This AI technology has the potential to enable more compact, lower cost, lighter weight optics and can be integrated into a variety of imaging platforms.
Improved low light performance.
Faster shutter speeds.
Reduced motion blur.
Motion or Still.
How ALLIS works
- Using any image sensor.
- A camera profile is learned and embedded into ALLIS, a form of AI based on machine learning.
- Using edge processing at a fraction of a second.
- The image produced results in:
- Low noise
- Improved low-light performance
- Faster shutter speeds.
- Reduced motion blur
How ALLIS Learns
- ALLIS learns from hundreds of thousands of example images
- Machine learning features are extracted from every image.
- ALLIS stores this information in its vast neural network.
- When a low-light high-noise image is captured on virtually any device…
- ALLIS applies its learning in a fraction of a second.
- Establishing a completed high-resolution low-noise image.
Security and IoT Devices
Reducing noise in images is a longstanding problem in the imaging industry, and an active area of research and development. Traditionally denoising is accomplished through hand-crafted algorithms or mathematical models that exploit the statistical underpinnings of the system’s noise. These methods can range from simple blurring of fine details, to costly techniques that require specialized hardware.
Regardless of their complexity these methods are limited by the assumptions an engineer makes about the system, and are created in isolation from the larger applications being created in the exciting field of computer vision.
Modern Low Light Modes
Smartphones have revolutionized the way we take photos. They’ve also accelerated the advancement of imaging software. Low light modes have become a must-have feature for taking photos on smartphones.
However, most low light modes aren’t suitable for video. They rely on taking multiple short exposures and combining them to create a single, noise free image. By comparison, NexOptic’s ALLIS is suitable for images and video.
ALLIS is a modern solution to the classic problem of imaging in low light. As a software solution, ALLIS is ready for licensing to a number of verticals relying of camera technology and has been designed to work “at the edge”.