search

experiencor / image-to-3d-bbox

Build a CNN network to predict 3D bounding box of car from 2D image.
https://experiencor.github.io/sdc_3d.html
237 stars 59 forks source link

Question for two files！ #12

Open cx2214496 opened 6 years ago

cx2214496 commented 6 years ago

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 commented 5 years ago

@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