AI super resolution demo uses Imagination accelerator
Visidon in Finland is using the latest neural network processor designs from Imagination Technologies for high resolution real time imaging in embedded applications.
The technology uses neural networks to low-resolution images and videos to upscale up to 4K and 8K. These can be implemented in real time in tight power and thermal SoC budgets using Imagination’s IMG Series4 neural network accelerator (NNA).
The NNA uses Tensor Tiling technology that allows Visidon to use AI enhancement for applications ranging from image restoration and colourisation, to drastically reducing video streaming bandwidth by up-sampling low resolution source content.
Related NNA articles
- Neural network core is optimised for robotaxis
- Graphics processing for AI in security, monitoring and more
- Imagination launches four RISC-V processor IP families
“Power efficiency and high performance are two of the main pillars that Visidon builds its solutions on. We were surprised to see our deep-learning networks run so easily on Imagination’s IMG Series4 NNA and we were impressed with the overall power efficiency maintained while processing high compute workloads,” said Matti Niskanen, CTO of Visidon in Oulu.
Visidon has developed a set of deep-learning-based super resolution networks that can upscale and remaster video streams up to 4K (2160p) and 8K (4320p) to fit modern display resolutions, increasing the video frame size while maintaining image details. Visidon often combines super resolution with noise reduction technology to eliminate noise increase while preserving real details and clarity.
More on Tensor Tiling and super resolution
Other articles on eeNews Europe
- Kontron set to launch Raspberry Pi4 system
- ST to launch its first neural microcontroller with NPU
- Semiconductor market heads for severe downturn says analyst
- FreeRTOS boost with secure microcontroller deals
- Siemens to discontinue business in Russia
