Any advice on getting the a.i. to generate an avatar with the whole body and head included in the picture. Im trying to get it to produce some designs i can use but most of the time the feature are cut off by the borders of the portrait.
Iâve had some luck with âdetailed whole bodyââŚstill get some that arenât complete but mostly the whole body comes throughâŚ
1 Like
I think Iâve seen someone suggest âzoomed out.â Donât know how well it works or if it does.
The problem here is that Glowforge is using a fairly outdated version of AI image generation models. This is something thatâs been solved in newer generations of the tech, but Glowforge hasnât admitted exactly which engine theyâre using so we canât even evaluate if an upgrade is on the horizon or not. This isnât like the rest of glowforgeâs software, where it can stay in one place and not feel completely out of date for a long while. These models are changing so quickly that youâll get left behind within three to six months if you try to bring an image generator to market like this.
The smart money is that they chose a system that they can swap out when newer generations come along, but newer generations are already here (and so much better), so Iâm not sure what the holdup is.
The specific issue in this case is that the these large image models generally crop all of the images they absorb into squares before processing them, which leads the model to believe that a lot of pictures of people have the heads clipped off at the top.
In the meantime, there are some tricks. Try using âwide angleâ or âportrait", sometimes that will skew the model toward looking at images that were not cropped when evaluating. You might also try âfrom Instagramâ, believe it not. Instagram has required a square image from the beginning so the images it pulled from there shouldnât be cropped.
Itâs a long shot but maybe worth giving it a chance.
The latest iteration of stable diffusion, 2.0, has largely fixed the cropped portrait issue, as has MidJourney, and probably some of the other more active players in this space. Pure speculation, but magic canvas feels like itâs based on stable diffusion 1.5, just judging by its overall look and feel. Weâll see if GF tries to keep current.
2 Likes
I too have had a look at the Magic Canvas and have come to the same conclusion. There are far more images that I canât use than there are those I can. Either the heads are completely missing or the image looks too weird (especially people). I keep telling myself that surely glowforge would correct this before launching it? Shame, I was really excited by the prospect of image help
Itâs improved a lot since the initial beta testing, so Iâm optimistic that it will continue to improve. 
1 Like
This topic was automatically closed 32 days after the last reply. New replies are no longer allowed.