MH/Valentina hackathon at Libre Graphics 2015

[slspencer]

MakeHuman developers interested in developing MH/Valentina 3D are invited to meet for a hackathon at
Libre Graphics Meeting 2015, http://libregraphicsmeeting.org/2015/ at the University of Toronto from April 29 - May 2.

FYI, last year's LGM was at the University of Leipzig. http://libregraphicsmeeting.org/2014/

If anyone would like to submit presentations on their work with MakeHuman please submit an application here:
http://libregraphicsmeeting.org/2015/call-for-participation/

It would be great to have MakeHuman developers at LGM2015.

[joepal]

That sounds very interesting. I'll repost it on the dev list to make sure people see it.

The cost of getting there and having some place to stay might be somewhat prohibitive for a project without a budget though. (for anyone who's interested, I ended up on around 1000-1100 euro when adding plane tickets from europe and hotel rooms).

[slspencer]

Libre Graphic's presenter application has fields to indicate if you need reimbursement for travel expenses.
Submit an application for a 20 minute presentation, and participation in the MH/Valentina Hackathon,
and check the box that you need reiumbursement.

[brkurt]

Libre Graphic's presenter application has fields to indicate if you need reimbursement for travel expenses.
Submit an application for a 20 minute presentation, and participation in the MH/Valentina Hackathon,
and check the box that you need reiumbursement.

I'm only a contributing artist to MH, and I am already fully (very!) employed, so I would like to pose an important question for you before you go to the hackathon.

Do you have a roadmap for Valentina development, and what commercial apps do you see as your comparison and/or competition?

If I remember my first Valentina exchange, there is a fashion software firm out of Tel Aviv that you saw as your main competitor, and it was I who alerted you to Marvelous Designer.

I say these things because I believe you will get more support from the ever-growing MakeHuman (and open-source) community if you present them with a road map, and then an engineering 'blue book'.

[slspencer]

Browzwear is the proprietary software that works similarly to where we want Valentina to go, with the difference being of course we wish to use real measurements from real individuals, as this is the base problem faced by the garment industry and is where we've made the breakthrough. Like Browzwear, Marvellous Designer can't be used to make garments for any particular individual IRL.

...what commercial apps do you see as your comparison and/or competition?

If I remember my first Valentina exchange, there is a fashion software firm out of Tel Aviv that you saw as your main competitor, and it was I who alerted you to Marvelous Designer.

[slspencer]

[quote="brkurt"]
Do you have a roadmap for Valentina development...

Valentina is currently aiming to become a minimally viable product.
We've solved the 2D pattern generation problem, and Grafis is the only real competitor, however there are other companies which partially implement custom-sized patternmaking tools.

The current roadmap is -
-Produce a Wavefront obj pattern using quadrangles
-Produce a 3D avatar using the same measurement set used to produce the pattern
-Capture 3D body scans and convert to mesh, then pull measurements from mesh to generate 3D avatar (Currently to develop the body scan module we are working with the Google Project Tango SDK. We will also work with Intel's Realsense SDK when it becomes available.)
-Import both pattern & avatar into Blender and stitch the pattern together into a garment around the avatar

Future modules:
-2D & 3D tools to aid in the design of garments to make the process more 'visual'
-Collaborative tools to enable work between designers and design teams
-Tools to select different fabrics, textures, notions etc. on displayed garments
-Tools to enable designers to order fabrics, notions, etc. from vendors
-Tools to enable interaction with manufacturers

The code base and issue list is here: bitbucket.org/dismine/valentina
Development & beta testing discussion is here: https://groups.google.com/forum/#!forum ... oject-list

[duststorm]

That's very interesting.
At MakeHuman we are also very interested in creating avatars from real measurements, be it from textual input or 3D scans. It's quite a big task though, this is where we could use extra helping hands.
I believe this is where Valentina and Makehuman can achieve great benefits from cooperating.

Lately I've been focusing on 3D scanning and was involved in the building of scanning installations for 3D capturing of people. At MakeHuman we are experimenting how we can use scanned content for producing qualitative 3D assets faster. These subjects are all very related, so if there are people at Valentina project focusing on the 3D scanning area, perhaps we could exchange thoughts and discuss technical approaches.
Do you guys have a Tango prototype device?

[slspencer]

Yes we have the Google Project Tango SDK.
It sounds like you're looking for people in the same areas we're looking for people.
All Valentina needs is to get people on board who can do the work, and we can get lots & lots of funding.
Our question is, why are 3D open source people so shy?
We don't have this much problem finding C++ and vector graphics developers.

That's very interesting.
At MakeHuman we are also very interested in creating avatars from real measurements, be it from textual input or 3D scans. It's quite a big task though, this is where we could use extra helping hands.
I believe this is where Valentina and Makehuman can achieve great benefits from cooperating.

Lately I've been focusing on 3D scanning and was involved in the building of scanning installations for 3D capturing of people. At MakeHuman we are experimenting how we can use scanned content for producing qualitative 3D assets faster. These subjects are all very related, so if there are people at Valentina project focusing on the 3D scanning area, perhaps we could exchange thoughts and discuss technical approaches.
Do you guys have a Tango prototype device?

[brkurt]

Yes we have the Google Project Tango SDK.
It sounds like you're looking for people in the same areas we're looking for people.
All Valentina needs is to get people on board who can do the work, and we can get lots & lots of funding.
Our question is, why are 3D open source people so shy?
We don't have this much problem finding C++ and vector graphics developers.

It may be due the size of the challenge that you have taken on:

(1) Must be an open-source solution, attractive enough to have a high number of developers;
(2) There must be a smooth flow from 2d patterns to 3d modelling (for the fittings), and then a complete draping/folding solution for every type of human;
(3) You need to account for all of the seven major types of folds (see Barbara Bradley's fold tutorial for this one).

So far--you can check my posts on this matter--Blender 2.7x can drape well, but not fold well; Marvelous Designer 4 both drapes and folds well, but the exported .obj is a mix of triangles and quads.

The Valentina solution can have no triangles, period. That will guarantee a loyal army of users and developers. Any triangles in the garment mesh, and DAZ, Maya, Modo (and Blender) users won't touch it. Especially the professionals, whose career is at stake.

[slspencer]

The Valentina solution can have no triangles, period.

We produced a 3d pattern with triangles because we're newbies to 3D.