gitouni / CalibNet_pytorch

Pytorch implementation of CalibNet
MIT License
31 stars 8 forks source link

A question about INFO - 2023-10-10 21:01:41,619 - test - [03201|05263],mdx:0.0624. what is mdx? #9

Closed cchester25 closed 4 months ago

gitouni commented 11 months ago

Please refer to https://github.com/gitouni/CalibNet_pytorch/blob/25fdfa519ea0963d80ad76a0817af886cb144442/test.py#L67-L71 It is the mean of the defined residual vector res_npy.

cchester25 commented 11 months ago

Please refer to

https://github.com/gitouni/CalibNet_pytorch/blob/25fdfa519ea0963d80ad76a0817af886cb144442/test.py#L67-L71

It is the mean of the defined residual vector res_npy.

Thanks for the answer, I would like to ask why the sequences 11, 12 and 13 must be present at the same time when testing, can't we just test a particular sequence?

gitouni commented 11 months ago

Please refer to https://github.com/gitouni/CalibNet_pytorch/blob/25fdfa519ea0963d80ad76a0817af886cb144442/test.py#L67-L71

It is the mean of the defined residual vector res_npy.

Thanks for the answer, I would like to ask why the sequences 11, 12 and 13 must be present at the same time when testing, can't we just test a particular sequence?

The sequence number can be designated in config.yml: https://github.com/gitouni/CalibNet_pytorch/blob/25fdfa519ea0963d80ad76a0817af886cb144442/config.yml#L1-L4

For example, to test CalibNet on only sequence 11, you just need to change the test parameter to [11].