GraphCMR

GraphCMR copied to clipboard

Thank you for your great works. I wonder what is the benefit of loading resnet50 pretrained? Why don't you training from scratch? If I load resnet50 pretrained, fix the parameters...

thaithanhtuan

Praise from a newbie

1

The code style and architecture are amazing! And I enjoy a lot. Thank you for sharing the neat code base.

tinatiansjz

Hi, In function [visualize_reconstruction()](https://github.com/nkolot/GraphCMR/blob/4e57dca4e9da305df99383ea6312e2b3de78c321/utils/renderer.py#L208), the default focal length is 1000., while in your [config file](https://github.com/nkolot/GraphCMR/blob/4e57dca4e9da305df99383ea6312e2b3de78c321/config.py#L88) and function [render()](https://github.com/nkolot/GraphCMR/blob/4e57dca4e9da305df99383ea6312e2b3de78c321/utils/renderer.py#L254), the default focal length is 5000. The function visualize_reconstruction() is only called...

krumo

I wonder why the SMPLParamRegressor don't go in an iterative way to give the final smpl param? Like described in paper End-to-end Recovery of Human Shape and Pose? I mean...

MengXinChengXuYuan

Compute 'A', 'D', 'U' matrices

6

Hello! I want to try use other human template(SMPLX) in your work, I need precompuute 'A', 'D', 'U' matrices according [this](https://github.com/anuragranj/coma) , so I test using the smpl template first,...

moondabaojian