gulvarol / bodynet

BodyNet: Volumetric Inference of 3D Human Body Shapes, ECCV 2018
http://www.di.ens.fr/willow/research/bodynet
MIT License
261 stars 42 forks source link

Want to generate the rotating gif of joe cole, how do we do this? #7

Open mezbot opened 5 years ago

mezbot commented 5 years ago

Hi @gulvarol I would like to create 3d model you have on the front of the repo, (rotating joe cole gif). How can I do this, what scripts need to be run? For clarity, we have run bodynet using pretrained weights, however when doing so our 3d voxel representation of joe cole does not look the same as your one on the README of the repo.

Any guidance here would be greatly appreciated

mezbot commented 5 years ago

@sheriefkhorshid

sheriefkhorshid commented 5 years ago

Hi @gulvarol When running demo.m , I managed to get all the images to render except for the "parts" image fig_parts(partclass, fn, vws);

seems to be using way too much memory

Error using smoothpatch_curvature_double Requested 140432545678274x1 (1046304.0GB) array exceeds maximum array size preference. Creation of arrays greater than this limit may take a long time and cause MATLAB to become unresponsive. See array size limit or preference panel for more information.

Error in smoothpatch (line 83) [Vx,Vy,Vz]=smoothpatch_curvature_double(double(FV.faces(:,1)),double(FV.faces(:,2)),double(FV.faces(:,3)),double(FV.vertices(:,1)),double(FV.vertices(:,2)),double(FV.vertices(:,3)),ceil(double(itt)),[double(lambda) double(sigma)],Ne);

Error in fig_parts (line 14) fv{pp} = smoothpatch(fv{pp}, 1, 5);

gulvarol commented 5 years ago

@mezbot you mean the script to rotate? Here is what I used once plotting the mesh with fig_mesh:

set(gca,'CameraViewAngle', 10)
for vv = 0:10:360
    view(vv, 0);
    drawnow;
end

The released model used in the experiments might not be the exact same epoch used to produce this particular visualization, but the visual result looks quite similar.

@sheriefkhorshid sorry I cannot reproduce your problem, it should not request that much memory.