the reason behind the usage of render_onehot_heatmap for training target instead of render_gaussian_heatmap

khanhha commented 3 years ago

Hi, I would like to ask about the difference between the two kinds of heatmaps in two repositories from you.

In the PoseFix repo, I noticed that you used render_one_heatmap for generating the training target, as shown below.

However, in another repo from you TF-SimpleHumanPose, the gaussian heatmap function is used for generating the training target instead, as shown below.

Could you please help me clarify two things?

Is there are paper/or description behind your render_one_hot_heatmap function?
Is there any advantage of your render_onehot_heatmap over render_gaussian_heatmap that motivated you to use it in your PoseFix repo?

Thank you for your patience and I look forward to hearing from you soon.

mks0601 commented 3 years ago

You can find answers of both questions in my PoseFix paper.

khanhha commented 3 years ago

Sorry for wasting your time. I should have read the paper more carefully! Best.

mks0601 commented 3 years ago

No problem! Let me know if you can't find them

mks0601 / PoseFix_RELEASE