We’re pleased to share that our paper, “Quality Consistency Score (QCS): A Survival-Based Reliability Descriptor for Video Quality Assessment,” has been accepted to IEEE ICIP 2026, taking place 13–17 September in Tampere, Finland.

Danish video enhancement company Pixop integrates with TVU MediaMesh, enabling broadcasters and streamers to deliver consistent, high-quality live video from any source without any infrastructure changes being required

Pixop co-founder and CTO Jon Frydensbjerg will join a panel of industry practitioners at the NAB Show 2026 Broadcast Engineering and IT (BEIT) Conference to discuss how machine learning is moving from experimentation to measurable impact across video workflows.

After discussing where Pixop came from and how it works, I wanted to understand why it exists. Why does a team of developers in Denmark spend years building filters that make video look better? And what’s next? So I asked Jon.

In Part 1, we followed Jon Frydensbjerg from his early Amiga days to his current work as Pixop’s co-founder and CTO. Now let’s dig into what Pixop actually does, and how its AI-powered filters work.

For the truly non-geek: This piece is aimed at non-technical audiences, too, and I’m told not everyone knows what a “sprite” is. It could be (1) a small magical creature, (2) a tiny, delicate person, (3) a two-dimensional image in computer graphics, (4) a large electrical discharge above thunderstorms, or (5) a lime-flavoured soft drink. Jon usually talks about option 3, but honestly, he brings a bit of all five. Oh, and “real-time” means that computers can do their magic on a video stream without having to delay or slow it (ie, it takes less than an hour to upscale an hour of video – In live scenarios, the magic happens in milliseconds.)