This isn't even redeemable from a technical perspective as Disney has been doing this same thing on their cruise ships -in the Animator's Palate dining hall- for at least a decade.
> This isn't even redeemable from a technical perspective as Disney has been doing this same thing on their cruise ships -in the Animator's Palate dining hall- for at least a decade.
The main interesting part about Facebook's approach is inferring bones from a humanoid drawing (even if other unrelated objects are on the page). From what I can tell off of some Youtube videos, the Disney version has guests draw the bones already separated on a template like this: https://i.imgur.com/AGxvsFq.png
A couple more things Facebook's method appears to handle better, but wouldn't be strictly out-of-reach for traditional methods, are morphing the limbs rather than having them as separate detached parts and identifying which parts of the background are inside of the figure so should be kept opaque. Comparison: https://i.imgur.com/KlJbtqy.png
Yes, true. I should have chosen my words more carefully. While the technology is better in a number of ways and the result is incrementally better in this specific application, the incrementally better result is arguably not worth the technical lengths and costs it took to get it. And what is not redeemable is the enticement, see here: https://imgur.com/vCyJgRE
The technically interesting thing is the automatic generation of the "bones", something the Disney solution does not do - they make you draw your character within a predefined grid.
But it's too bad that every new piece of technology nowadays has a huge asterisk attached: please sign away your family's privacy and security, We Promise You Can Trust Us. I guess I'll just have to pass.
The main interesting part about Facebook's approach is inferring bones from a humanoid drawing (even if other unrelated objects are on the page). From what I can tell off of some Youtube videos, the Disney version has guests draw the bones already separated on a template like this: https://i.imgur.com/AGxvsFq.png
A couple more things Facebook's method appears to handle better, but wouldn't be strictly out-of-reach for traditional methods, are morphing the limbs rather than having them as separate detached parts and identifying which parts of the background are inside of the figure so should be kept opaque. Comparison: https://i.imgur.com/KlJbtqy.png