[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k] [s4s] [vip] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Settings Home
/3/ - 3DCG

Thread archived.
You cannot reply anymore.

File: 15056838201641.gif (2.49 MB, 316x322)
2.49 MB
2.49 MB GIF

Very good!

File: Clipboard.jpg (60 KB, 1254x532)
60 KB
It's neat and clever but I don't know about very good.
The one semi-neat thing is that the .OBJ files it exports are vertex painted to retain the original photo (more or less). This unfortunately requires your importer to support .OBJ vertex paints which last I checked isn't actually part of the Wavefront spec but an unauthorized extension (Meshlab supports it).
File: Untitled.jpg (29 KB, 257x467)
29 KB
File: doubting too hard.jpg (33 KB, 322x454)
33 KB
File: alexwants.gif (492 KB, 158x187)
492 KB
492 KB GIF
Lmao, this site has potential

what have we done
File: ah-knee-may.png (129 KB, 462x466)
129 KB
129 KB PNG
do me a solid and remind me what the source image is
File: ah-knee-may_source.jpg (29 KB, 320x320)
29 KB
pic related is the picture i used
the source is by an artist named apostle
you can see the full pic at https://chan.sankakucomplex.com/post/show/5569986
> there will soon be 3D Shadmans
Hell is full
>>there will soon be 3D Shadmans
there already is. skuddbutt regularly makes 3d modelled porn of and for shadman and people affiliated with shadman (namely jlullaby)
lewd, reported.
File: kono smile.jpg (94 KB, 545x764)
94 KB
>you could pretty much use this to output the OBJ and have a pretty easy template to trace for modelling
This guy gets it

Did it fit?
I expected a lot worse honestly
File: Capture.jpg (21 KB, 453x479)
21 KB
ay bby
Full body when ??
Anyone else gonna email this guy with useful information?

I'm convinced he can take this into a sort of Facegen route and try to conform a preset topology that can dynamically alter its edgeloops and faceloops to match the level of detail it can confidently extract and end up with a better result because it would be working with adjusting musculature to the form, instead of just this raw sort of voxelized 3D volume.
File: 222222.png (80 KB, 126x264)
80 KB
reminds me more of a bump mapping with face recognition
File: mene-man.jpg (18 KB, 438x462)
18 KB
File: Clipboard.jpg (126 KB, 1563x602)
126 KB
126 KB JPG
It seems that this was trained on 3D scans from mobile phones so the results are about what you would expect from a mobile phone 3D scan.

But what if he taught it based on one Daz model and all of its possible morphs and sliders?

I imagine at the very least the topology would look better, but I don't know if it'd handle real-world photographs well.

Anyway I thought it would be interesting to see how close it could get to a Daz model I already have. It has the general mass but it's not even worth retopologizing or sculpting.

https://www.daz3d.com/fwsa-adaline-hd-for-victoria-7 is the store page.
File: geaeags7tk.jpg (354 KB, 640x960)
354 KB
354 KB JPG
This is fun. I'm sure the novelty will wear off though.
The models are pretty rudimentary not to mention messy.
File: yfcjcgk.jpg (143 KB, 640x480)
143 KB
143 KB JPG
Whoops, forgot the image.
true, but it would still be great for shitpost animations along the lines of mr muskel
File: Cruise face.webm (1.92 MB, 720x486)
1.92 MB
1.92 MB WEBM
Here's something for your shitposting animations.
thanks, i didn't notice that

was able to export a DAE and i can import that with vertex colors just fine
What's the easiest way to get the vertex colors into 3ds?
don't know anything about 3DS but chances are it supports DAE and FBX, both formats have vertex color support. use meshlab http://www.meshlab.net/ to import the OBJs and then go to export as and pick DAE or FBX and try those in 3DS

Hail Xenu!
Here's a video explanation of the site by the way: https://www.youtube.com/watch?v=JtK4cTLlUko
Some of the smartest people out there... and they can't level the audio volume.

Still neat.
Aw man the model he showed off didn't have the weird blocky lips and that's apparently what it's actually capable of but the demo is "compressed" a bit to be easier to demo for everyone.
Well the code and the paper on it is up on the site for everyone to take advantage of.
You could always run it on your own computer.
lol I don't think so bud, I already got a space heater!
Yeah but this is a face heater. :^)
File: 20170920_193642.png (172 KB, 531x610)
172 KB
172 KB PNG
mm, yes
File: 20170920_193741.png (88 KB, 466x682)
88 KB
very good indeed
File: 1.jpg (46 KB, 500x500)
46 KB
wait a minute is this just like crazy bump making a fucking displacement out of an image?
No, Crazy Bump uses how bright or dark areas of an image are to make a normal map.

This, for a tl;dr, knows what collections of pixels correlate to what 3D volumes in past situations and estimates the volume of faces in new situations based on those memories.

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.