cassie101

reg label 生成

1

https://github.com/fire717/movenet.pytorch/blob/bbc81408bd4da49789d912fd08635355fe123e60/lib/data/data_tools.py#L123-L136 您好，感谢您的分享，看了这段关于reg生成heatmap的代码，是在cx，cy周围（左与上边2个pixel；右与下边3个pixel）也附上reg值,但在不同位置时调整偏移是根据 `img_size//4/2-1`，相当与图的一半。这样的操作导致比如 - cx,cy在图上半部分，reg值在cx,cy周围反而偏多或少了?

Why *0.5 the loss?

2

Hi @bearpaw thank you for sharing the work, may I know what does the 0.5 stand for in the loss function? `loss += 0.5 * self.criterion(heatmap_pred, heatmap_gt)`

Hi @wywu, we noticed that some of the points for testing are strangely labeled. For example, in image 20_Family_Group_Family_Group_20_118.jpg, the landmark for point 88 (left inner lips) and 92 (right...

codeserver on sagemaker studio

9

Hi, codeserver seems to be running bare bone, similar to a system terminal. Is there anywhere we could run codeserver that uses computing resource with gpu such as pytorch-1-8-gpu-py3-ml-g4dn-xlarge?