search

Ninebell / GaneratedHandsForReal_TIME

it referenced paper of GANerated Hands for Real-Time 3D Hand Tracking from Monocular RGB.
MIT License
86 stars 17 forks source link

Weight of Silnet and GeoConGan #7

Closed hizircanbayram closed 4 years ago

hizircanbayram commented 4 years ago

Hello! I was wondering how we can access weights of Silnet and GeoConGan. Is there any trained model that you can share with us? If there was not, can we get those weight training your projects? If so, how? Thanks in advance.

Ninebell commented 4 years ago

hello, if you tell me your email, I will send you

hizircanbayram commented 4 years ago

Thanks! hizircanbayram@gmail.com

TingHungEricLin commented 4 years ago

Dear Ninebell,

Would I get, too? ericthlin@gmail.com

TingHungEricLin commented 4 years ago

Dear Ninebell,

Would you please re-send this to me? I made mistake, so you sending email (for GeoConGan) delete on google clouding... Please, and thanks...

TingHungEricLin commented 4 years ago

For 64 64, I did change heatmaps_pred3D = RenderingLayer([64,64], coeff=1, name='heatmapspred3D')(projLayer) -- regnet.py & self.base = k_b.ones((1, 6464), dtype=np.float) -- ProjLayer.py...

But, get a wrong messages... concat_pred_rendered = concatenate([conv, conv_rendered_3]) ... A Concatenate layer requires inputs with matching shapes except for the concat axis. Got inputs shapes: [(None, 8, 8, 256), (None, 16, 16, 128)]

Do I miss to change another code?

xiaoguitoucode commented 4 years ago

Hi, Could you send me the weights of Silnet and GeoConGAN? Thanks ! 1130455704@qq.com