Visidon AI-powered Low-Light Video Enhancement selected to Hailo-15 AI Vision Processor
The CNN-based technology developed by Visidon allows for significant improvement to video analytics accuracy in low-light environments, marking a new era for intelligent cameras deployed in public spaces, smart cities, factories, buildings, retail locations, and more.
March 12th 2024
The leading AI chipmaker Hailo has selected Visidon CNN-powered low-light video enhancement for their Hailo-15 AI vision processor. Hailo-15 is a family of AI vision processors for smart cameras that deliver up to 20 TOPS of AI inference and are able to process deep-learning AI applications such as video analytics.
By introducing superior AI capabilities into the camera, Hailo is addressing the growing demand in the market for enhanced video processing and analytic capabilities at the edge. With this unparalleled AI capacity, Hailo-15-empowered cameras can carry out both video enhancement and significantly more video analytics, running several AI tasks in parallel including faster detection at high resolution to enable identification of smaller and more distant objects with higher accuracy and less false alarms.
Founded in 2006, Visidon specializes in developing AI-based image and video enhancement technologies. They have developed numerous algorithms tailored for embedded imaging, collaborating closely with embedded camera vendors, such as mobile OEMs. Thanks to their extensive experience in embedded imaging, Visidon was able to offer the most competitive low-light enhancement in the market. The technology includes state-of-the-art noise reduction algorithms, ensuring clear and crisp images even in challenging lighting conditions. Noise reduction enhances the quality of captured footage, improving the accuracy of video analytics and enabling better decision-making in various applications. Through optimization of the network, Visidon efficiently utilizes the limited power assigned for low-light improvement, resulting in significantly superior outcomes compared to traditional ISP (image signal processor) technologies even in ultra-low-light, such as below 0.1 lux. Notably, Visidon's low-light enhancement performs well even when objects are in motion.
“We are excited for collaborating with Hailo to enable a remarkable low-light video quality for Hailo-15 AI vision processor-empowered camera devices with our AI de-noise technology. Not only for improving visual quality, but also to increase AI detection accuracy in challenging conditions offering a real competitive edge for Hailo-15 smart camera customers”, comments CEO of Visidon, Markus Turtinen.
Visidon’s solution is fully integrated into the Hailo-15 software stack and can process 4K stream at up to 60fps and in lighting conditions as low as 0.1 lux while maintaining color information, sharpness and producing minimal ghosting effects. Due to the high capacity of the Hailo-15 neural core, advanced AI analytics such as License plate recognition (LPR) are able to run in parallel with no reduction in performance.
“Our partnership with Visidon is based on a shared belief that the future of ISPs is going to be neural networks based. AI-driven image quality has become standard in smartphones, and we are looking to bring the same level of algorithmic and hardware innovation to smart-cameras. Our neural network (NN) core is unique in its ability to efficiently process 4K streams using minimal DDR bandwidth and power. By combining this with Visidon's proficiency in neural image enhancement, we've achieved truly remarkable results” said Mark Grobman, ML CTO at Hailo.
Visidon has a long history of collaboration with platform providers, and their low-light enhancement technology is hardware-independent, ensuring its compatibility across different systems. The technology will be next shown live at Embedded World in Nürnberg 9-11 April and ISC West in Las Vegas 10-12 April.
See a demo video of the solution: https://visidoncloud.com/s/8nCrrBZsTrzYr6A
Contact
Hailo
Visidon
Markus Turtinen
CEO
PR contact
Jenna Enbuska
Head of Marketing
Tags: