Skylum also hires Alex Savsunenko, former CEO of Let’s Enhance, to lead AI-based photo technologies to the next level.
Skylum Software announced the formation of a new research and development division dedicated to the advancement of artificial intelligence technologies in image processing. The Skylum AI Lab leverages the company’s prior work developing smart filters in its award-winning Luminar software, as well as technology from its “sister company” Photolemur, which was founded in 2016 by Dima Sytnik and Alex Tsepko, co-founder and CEO of Skylum respectively.
“Clearly, AI can simplify our lives. By using AI-based technologies in our products, our customers save time vs. manual editing, and can often get better results,” said Alex Tsepko, CEO at Skylum. “Our neural networks are being trained on millions of images taken by cameras from Fujifilm, Sony, Olympus, Nikon, Canon and many others, which means outstanding results for all photographers, regardless of what style they shoot and what gear they are using.”
To spearhead the new Skylum AI Lab, the company has hired Alex Savsunenko, former CEO of Let’s Enhance, a leader in machine learning for visual content. Savsunenko will manage all research and development for technologies based on artificial intelligence, machine learning, and neural networks. Promising results will ultimately be implemented in Skylum products and solutions for image and video enhancement, with the aim to help users make their workflow faster, smarter and more efficient.
Currently, the Skylum AI Lab is testing over a dozen of new solutions, including:
- Image upscaling: uses deep convolutional neural networks to improve low-resolution images and scale them up for superior viewing and printing.
- Tagging: generates tags that describe the image and its objects based on image recognition.
- Segmentation: smart recognition of image areas that can be automatically enhanced using different filters and corrections based on the type of object.
- Automatic enhancement: applies smart image corrections to photos based on a variety of issues