about testing

[River-Zhang]

Hi, I was wondering if you could provide the testing and training code so that we can test and train on other datasets! Thanks very much!

[YuliangXiu]

• testing code: it's quite messy, I will release the benchmark.py for evaluation after cleaning
• training code: please refer to #21

I highly recommend using ECON_EX for testing or evaluation, which is optimization-based and requires no training process. And the normal and SMPL-X estimators are borrowed from off-the-shelf approaches.

[River-Zhang]

Thanks very much！ By the way, I also want to ask something about the 3D results' images in the paper. Could you please tell me what rendering tools did you use to turn the mesh into such nice pictures?

[YuliangXiu]

Very simple

1. Meshlab snapshot (phong.gdb, max shininess)
2. Google Drawing (Format options -- Recolor)

Meshlab	Google Drawing

[YuliangXiu]

@River-Zhang Testing code is ready to use, please check testing.md for more details

[River-Zhang]

Copy that！ Thanks very much!