Question for two files！

[cx2214496]

Hi，@experiencor
Can you explain to me the function of these two codes? Does the execution of the second code depend on the execution of the first code?
Thanks man！

[lucasjinreal]

@cx2214496 The model is trained on kitti dataset, in kitii, object not only provide 2d box, but also object width, height, length and anchor information, which is possible to discribe position of a object in projection view.

That model is implemented by a paper: https://arxiv.org/pdf/1612.00496.pdf

And the output is as label: dimensions, orientations, and possiblities. This repo give u a way to make the prediction back to the 3d box boundries.

I am still training the model, after that, let me post some result about this