Hacker News Reader

I'm not sure I see value here. All it's doing is providing a UI for switching between keyframes. The UI doesn't strike me as particularly intuitive. The resulting animations aren't particularly good.

It looks like a fun demo, but what's the point?

It's generating all of those keyframes from a single picture. The slides are quite interesting: https://docs.google.com/presentation/d/11ia5Zeg7AiiIi5y10S6z...

I actually think this is a really amazing demo and am excited about the possibilities of this kind of "high efficiency" interface. I'm imagining it as a first step towards the next stage of machine-learning-powered workflow tools. This demo might even be the first step to higher efficiency character animation for game dev shops!

The tracing box as a UI is very interesting too. I haven't seen it used very often but it looks powerful and flexible. Keep up the great work!

This is awesome!!

I definitely want to play with this.

TopoSketch – Generating Animations by Sketching in Conceptual Space