Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Human Motion Diffusion Model

github.com

Post reply on HN
Print view View original

1–10 of 32 posts

Re: Human Motion Diffusion Model

#5

Post by cdev_gl »

soperj wrote: post #2

I understand it's probably gonna get better and better, but the actual result made me laugh out loud. The skipping rope one was hilarious.

If you scroll down a bit there's a wireframe of the skeleton which is what's actually being animated, and you'll notice it's lacking in bones to define the fingers or even possibly hands. Hence why the hands maintain that weird pose throughout all examples.

My gut says that the quality could be rapidly improved without changing the underlying design at all.

The real issue with this, I think, is that motion capture for humans is already widely available and provides much higher fidelity and control than text. Unless I'm misreading the paper badly, this model was trained on exactly such data. Blending between multiple animations through motion capture is also well-understood.

So while the results are impressive, the practical gains seem very marginal. I think perhaps that the equivalent to "inpainting" (as mentioned in the text) and "style transfer" would be the big gain here? If we could use this to retarget animations to different body plans (child, adult, space monster) quickly, or for smarter interpolation between human-authored keyframes, I could see that being a much-desired tool.

Re: Human Motion Diffusion Model

#6

Post by dagmx »

cdev_gl wrote: post #5
soperj wrote: post #2

I understand it's probably gonna get better and better, but the actual result made me laugh out loud. The skipping rope one was hilarious.

If you scroll down a bit there's a wireframe of the skeleton which is what's actually being animated, and you'll notice it's lacking in bones to define the fingers or even possibly hands. Hence why the hands maintain that weird pose throughout all examples. My gut says that the quality could be rapidly improved without changing the underlying design at all. The real issue with this, I think, is that motion capture fo…

The reason there’s likely no finger joints is because a lot of motion capture data doesn’t include fidelity beyond the wrist.

So if they’re training on the standard corpuses of motion capture data available and even mixing in their own, they likely won’t have fingers to base data on.

Re: Human Motion Diffusion Model

#10

Post by soperj »

Agentlien wrote: post #4
soperj wrote: post #2

I understand it's probably gonna get better and better, but the actual result made me laugh out loud. The skipping rope one was hilarious.

I was actually really impressed by these results. But yes, that skipping rope does look hilarious.

Oh it's definitely impressive, I just couldn't help laughing.
Post reply on HN