vcarlosrb / 3d-body-measurements Goto Github PK

From where can i get Person.obj file
File is produced from blender or by coding

Hello Carlos, thank you very much for the work you have done! Would it be possible to implement the shoulder circumference? Is it in your skills?

Thank you very much!

Hi
Can anyone tell me how to convert obj file from image ?

I come again.When I execute the getHeight method, I find that the return value is direct (maxY - minY) without conversion. This should calculate the height of the model, not the height of the person. How to determine the conversion ratio

Hello, first of all, thank you for your code to help me with anthropometry. In addition, I would like to know whether the measurement method at each location is documented or whether you have published a paper. I would like to refer to the specific implementation process

I am having this issue while running on a custom obj file

Hello, thank you for your amazing implementation
Can you guide me how can i get vertices and faces for my custom images to use with this implementation ?
It would be great if you can provide a Google Colab Notebook Demo for this repo, it'll help everyone and it'll be much appreciated,

Thank you

Hi,

any news about the weight validation. Is it still pending ?

I am very interested in this feature.

Hi
Thanks for the nice repo and library .
We tried generating a 3d model with expose https://github.com/vchoutas/expose which tries to fit a smplx body model .
After that we used your library to derive the measurements and here is the error .
Thanks
Unimplemented OBJ format statement 'g' on line 'g model' Unimplemented OBJ format statement 's' on line 's 1' Traceback (most recent call last): File "/home/ujjawal/my_work/Fashion/3d-body-measurements/example/example.py", line 29, in <module> main() File "/home/ujjawal/my_work/Fashion/3d-body-measurements/example/example.py", line 19, in main body = Body3D(vertices, faces) File "/home/ujjawal/miniconda2/envs/py3.9/lib/python3.9/site-packages/body_measurements/measurement.py", line 192, in __init__ _, self.hip_location, _ = getHip(self.sections, self.crotch_location) File "/home/ujjawal/miniconda2/envs/py3.9/lib/python3.9/site-packages/body_measurements/measurement.py", line 87, in getHip polygon = getLargerAreaPolygon(sections[cont]) TypeError: list indices must be integers or slices, not NoneType

Hello，Sorry to bother you! When I execute the getSections method, what does len (entities) in the return value mean?

I ran pip install. Could you please let me know how to set up my environment to try testing this repo. I'd greatly appreciate it.

Hi cameronking4. If you want to test the library on your machine without installing the dependency via pip install, I recommend you follow the following steps:

1. Create a python environment: python -m venv ENV
2. Activate created environment: source ENV/bin/activate
3. Install the following dependencies: trimesh> = 3.8.8, numpy> = 1.19.1, scipy> = 1.5.2, shapely> = 1.7.1 and networkx> = 2.5
4. Run the example inside the example folder: python example.py

I followed these steps but when I execute example.py, the terminal waits for a second and then nothing.
How do I see the output? What am I missing? Where am I supposed to be looking?

Originally posted by @jtorres122 in #5 (comment)

How to generate that vertices in .obj file from images ?

I am getting this error:

FileNotFoundError: Could not find module 'C:\Users...\miniconda3\Library\bin\geos_c.dll' (or one of its dependencies). Try using the full path with constructor syntax.

The shoulder length from the project is very inaccurate.

This is a SMPL model of a user. The real shoulder measurement is close to 40cm but the program outputs 102cm. Is there something wrong I'm doing?

I'm glad to see this project. I wonder if this project supports anthropometry? I need to measure the human parts of the generated manikin. I hope I can get your reply and information.

Hi @vcarlosrb ,
I'm getting incorrect measuremnts from the mesh.
Steps to generate my Apose model is:-

1. I have used your suggested repo to generate smpl model (https://github.com/gulvarol/smplpytorch)
2. I generated shape parameters from smplify-x repo results ('betas') (https://github.com/vchoutas/smplify-x)
3. Pose parameters is same as you suggested
   I can see the mesh is generated well in Apose.

Here is the code to generate my Apose body model:
`def create_Apose(parameters_filepath,gender,outfile):
cuda = True
batch_size = 1

# Create the SMPL layer
smpl_layer = SMPL_Layer(
    center_idx=0,
    gender=gender,
    model_root='/home/ujjawal/my_work/Fashion/3d-body-measurements/smplpytorch/smplpytorch/native/models')

# Generate random pose and shape parameters
#pose_params = torch.rand(batch_size, 72) * 0.2
#shape_params = torch.rand(batch_size, 10) * 0.03
pose_params=np.zeros(72,dtype='float32')
pose_params[50],pose_params[53]=5.6,-5.6
f=open(parameters_filepath,'rb')
mydict=pickle.load(f)
shape_params=mydict['betas'][0]
#b=np.load(parameters_filepath)
#shape_params=b['betas.npy']
print('shape params:',shape_params)
print(pose_params.shape)
print(shape_params.shape)


# GPU mode
if cuda:
    pose_params = torch.from_numpy(pose_params.reshape(1,pose_params.shape[0])).cuda()
    shape_params = torch.from_numpy(shape_params.reshape(1,shape_params.shape[0])).cuda()
    smpl_layer.cuda()
print(pose_params.shape)
print(shape_params.shape)

# Forward from the SMPL layer
verts, Jtr = smpl_layer(pose_params.float(), th_betas=shape_params.float())
mesh=trimesh.Trimesh(verts.cpu()[0],smpl_layer.th_faces.cpu())
#trimesh.exchange.obj.export_obj(mesh)
trimesh.exchange.export.export_mesh(mesh,outfile,file_type='obj')
#mesh.show()`

Can you please suggest me any proper way to generate mesh for these measurements?

Traceback (most recent call last):
File "D:\3d-body-measurements-main\3d-body-measurements-main\example\example.py", line 37, in
main()
File "D:\3d-body-measurements-main\3d-body-measurements-main\example\example.py", line 19, in main
body = Body3D(vertices, faces)
^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Admin\AppData\Local\Programs\Python\Python311\Lib\site-packages\body_measurements\measurement.py", line 212, in init
self.sections = getSections(self.mesh, self.levels)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Admin\AppData\Local\Programs\Python\Python311\Lib\site-packages\body_measurements\measurement.py", line 172, in getSections
sections = mesh.section_multiplane(plane_origin=mesh.centroid,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Admin\AppData\Local\Programs\Python\Python311\Lib\site-packages\trimesh\base.py", line 2264, in section_multiplane
lines, transforms, faces = intersections.mesh_multiplane(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Admin\AppData\Local\Programs\Python\Python311\Lib\site-packages\trimesh\intersections.py", line 237, in mesh_multiplane
vertex_dots = np.dot(plane_normal, (mesh.vertices - plane_origin).T)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<array_function internals>", line 200, in dot
ValueError: shapes (3,) and (6,45966) not aligned: 3 (dim 0) != 6 (dim 0)

With 50 and 53 pose params set to 5.6 and -5.6, the SMPL model gives pose as shown:

I am using smplx package with model_type='smpl'.

Hi Carlos, I hope you are doing well.

Question: is it possible to generate a STAR or SMPL body model based on the following input parameters?

• Age
• Sex
• Height
• Weight

Preseize.ai appears to be doing this, but I don't know how.

Thanks!