3D hand normalization with empty frames

[J22Melody]

I am trying the code from https://github.com/sign-language-processing/pose/pull/36.

When I input a .pose file where the first few frames are empty (=no right hand detected and all values are 0s), normalization output becomes all nans, where 0s are expected.

Since the pose data is a masked array, I use pose.body.data.toflex() to inspect the arrays and now I have to use np.nan_to_num(pose.body.data) as a workaround to convert the nans back to 0s.

@AmitMY maybe worth checking the 0-handing logic in the code?

[AmitMY]

I'll need to see a reproduction (google colab perhaps). In any case, the correct way to get the value from a masked array is pose.body.data.filled(0)

[J22Melody]

https://colab.research.google.com/drive/1XyfAoKeTDRakiCHYEFgsEoceciaIlsar

[AmitMY]

Now this is done in the code https://github.com/sign-language-processing/pose/commit/d98ca47aed3e48dd41b885f521b755163893e2fa