Machine Learning used to upscale a Lumière film

February 2020: YouTuber Denis Shiryaev has published an « upscaled and resounded » 60 FPS version of the classic Arrival of a Train at La Ciotat (1897).

The resulting video:

According to Ars Technica:

How did Shiryaev do it? He says he used commercial image-editing software called Gigapixel AI. Created by Topaz Labs, the package allows customers to upscale images by up to 600 percent. Using sophisticated neural networks, Gigapixel AI adds realistic details into an image to avoid making it look blurry as it’s scaled up.

Petapixel writes:

To achieve this, Shiryaev used two programs: Google’s DAIN and Topaz Labs’ Gigapixel AI. Gigapixel AI was used to upscale the puny resolution video all the way to 4K, while DAIN was used to create and insert frames where they didn’t exist before, increasing the FPS of the video to 60.

In the words of Denis Shiryaev:

In some articles they are crediting me for having done something unique, but in my opinion this is unfair. Anyone can repeat this process with algorithms that are currently published on Github; all of them are in the video description. Credit should go to DIAN, Topaz AI, ESRGAN, Waifu2x, DeOldify, Anime 4K and other developers who are part of the worldwide ML-community and contributing to humanity by making these algorithms publicly available.

Thanks to Sebastian Bohle for bringing this to my attention!