Artificial Intelligence: AI Inside The New iPad Will Boost Low-Resolution Pics
It’s possible with the Apple Neural Engine, iPad’s dedicated neural processor.
The ML Super Resolution is a new tool inside Pixelmator Photo 1.4, a powerful photo editor created exclusively for the iPad. The new tool increases the size and resolution of images intelligently and automatically. This functionality, which was available in Macs, now comes to the iPad Air. (Korea IT Times)
iPad processing power
Pixelmator Photo 1.4 has been developed by Pixelmator Team, a Lithuanian company. It presented the ML Super Resolution at this year’s “Time Flies” Apple (NASDAQ: AAPL) event.
ML Super Resolution will have the distinction of being the first AI-powered image enhancement tool available on a mobile device.
That’s thanks to the new grunt available with the iPad’s processor. The new A14 Bionic chip in the latest iPad Air has some serious processing power. It can therefore take on whatever the ML Super Resolution can throw at it – and that’s saying something – because the AI feature needs up to 62,000 times more processing power than traditional approaches.
“The processing power of iPad has advanced in leaps and bounds over the last few years,” said Tomas Andrijauskas, the lead developer of Pixelmator Photo, to Apple Insider. “With these advances, it’s now possible to open up workflows that simply weren’t possible in the past, like using machine learning techniques to magically enlarge photos while retaining sharpness and enhancing intricate details.”
So if you have small, tightly cropped photos, or images that were taken with older cameras, no worries. With a tap, the algorithm inside Pixelmator Photo will enlarge the image while preserving and even enhancing textures, edges, and other details.
How it works
ML Super Resolution first creates a layered representation of the image that is over 100 channels deep. It then uses a Core ML-powered algorithm to detect features such as edges, patterns, textures, gradients, and colors.
It then separates them and upscales each type of feature separately, before combining them back into an image.
Related Story: Apple Acquires Inductiv Inc., a Machine Learning Startup
Image: Pixelmator Blog
Latest Alternative Investment News
The Brunel Pension Partnership has picked Ownership Capital, RBC Global Asset Management, and Nordea Asset Management to manage its new Sustainable Equities Fund of around £1.2 billion ($1.55 billion). Brunel…
CrowdProperty, a leading crowdfunder/lender to the U.K. SME property sector, has lent over £100 million to development professionals. The funding had a beneficial impact on the UK economy because it…
Ripple Labs, the San Francisco-based technology company that facilitates cross-border payments through the use of its blockchain solutions via RippleNet, will invest an undisclosed amount in Money Tap. Money Tap…
Earlier this year, Apple (NASDAQ: AAPL) acquired Vilynx, a Barcelona-based startup that uses AI and computer vision technology to analyze a video’s visual, audio, and text elements. Reportedly, the iPhone…