anilbas / 3dmmasstn Goto Github PK

I want know the detailed way to obtain the BFM_UV.mat. Could you tell me the method? thanks.

hi @anilbas,

Thanks for sharing the code!

I followed your readme to run the demo.m, but i can't find vl_nnbilinearsampler function. Could you please release that part of code?

sampler = vl_nnbilinearsampler(Images,grid);
demo (line 131)

Thanks in advance!

hi @anilbas,

Thanks for sharing the code!

I followed your readme to run the code, but there seemed to be some problems. Could you please give me some advice on the questions below?
In dagnn_3dmmasstn.m line 39, load('model/model.mat'), what should the model.mat be?
Besides, what is the training set, and how to create the imdb.mat?

Thanks in advance!

Thanks for your work and codes.
I can't find the imdb.mat file in this demo, can you provide the download link or tell me how to generate this file?
Thanks~

hi @anilbas,
Thank you for your contribution!
I followed your readme to run the code,but I can't run the training code.Could you please give me some advice on the questions below?
In dagnn_3dmmasstn_init.m line 7. This xavier(1,1,4096,16) should be the algorithm that initializes the network weight, but I run the prompt without definition.
Thanks in advance!

hi @anilbas,
Thank you for your great work.
I'd like to try some images, but can not find your model.
Where can I get your model for test?
Thank you!

I have read your paper but I don't understand ‘ computing the Tutte embedding with conformal Laplacian weights for UV graph' that you mentioned in your paper .Could you kindly share the code ? Or just give me some advice about it ?And I try to use matlab's Toolbox Graph,but my result is very different from yous

Thank you very much for your contribution to the academic society!

However, error occurred when a new layer was added to the net.
The code is:
net.layers(end+1) = struct( 'name', 'theta', ... 'type', 'conv', ... 'weights', {weightsandbias}, ... 'stride', 1, ... 'pad', 0, ... 'learningRate', opts.theta_learningRate) ;

Error information:

Subscripted assignment between dissimilar structure.

Error dagnn_3dmmasstn_init (line 18)
net.layers(end+1) = struct( 'name', 'theta', ...

Error dagnn_3dmmasstn (line 38)
net = dagnn_3dmmasstn_init(model,idx,opts);

The layer you want to insert does not have the same content.
The layers of the original net does not have 'weights' and 'learningRate', and the 'pad', 'stride' are in the 'block' item of these layers.

Hello,

Thanks for the awesome work! I'm a complete noob in the 3d modelling world so a lot of things fly over my head unfortunately. Please can you tell me how you produced "geometry.png". Is there some code i could look at to help me get that geometry image. I'm trying to produce it by looking at your current code but no luck.

Hi, Anil
What's your label of training data looks like? It is the same as the label in demodata.mat?

thanks very much