Segmentation Fault (core dumped)

[JasOlean]

Segmentation Fault (core dumped)

[lshiwjx]

Segmentation fault is too broad, you can try to debug it line by line.

[RayLeijingsi]

When I extract bone data file for kinetics, I got segmentation fault (core dumped) error for kinetics_train.
It is ok for kinetics_val. How to fix this?

I also met this problem when i used nturgbd-view_train. The problem is in the tools.py (AD=np.dot(Dn, A)). It cant handle more than 21 nodes. I don't know how to fix it. Have you fix it?

[JasOlean]

For kinetics, there are 18 joints. But I met this problem. Until now, I can't fix it.
By the way, when you train with num_person = 1, do you face "RuntimeError:"?

[RayLeijingsi]

For kinetics, there are 18 joints. B

on kinetics？

For kinetics, there are 18 joints. But I met this problem. Until now, I can't fix it.
By the way, when you train with num_person = 1, do you face "RuntimeError:"?

Sorry, I have not changed this parameter.
For segmentation fault, have you find the wrong place or the codes end place?

[JasOlean]

It didn't show anything except "kinetics train Segmentation fault (core
dumped)" after finish extracting kinetics val.

[JasOlean]

So, did you already extract train_data_bone.npy file for kinetics dataset. Please, could you send me this file to me?

[RayLeijingsi]

So, did you already extract train_data_bone.npy file for kinetics dataset. Please, could you send me this file to me?
i'm so sorry , I am restricted to send data because my computer can't be allowed to access the Internet.
But we can keep touch. this my email [email protected]