What are the camera calibration matrices used for in training?

maudzung / SFA3D

Super Fast and Accurate 3D Object Detection based on 3D LiDAR Point Clouds (The PyTorch implementation)

https://github.com/maudzung/Super-Fast-Accurate-3D-Object-Detection

MIT License

1k stars 271 forks source link

What are the camera calibration matrices used for in training? #65

Closed sisaha9 closed 2 years ago

sisaha9 commented 2 years ago

Can we make do without them?

maudzung commented 2 years ago

No, I used only lidar data to train the SFA3D model. Images were only used for output visualization.

sisaha9 commented 2 years ago

Hi @maudzung . Thanks for the quick reply. We tested by removing the images (which worked) and the camera calibration matrices (which did not work). So it is being read in for the train script. Would you know why/which file I should look at to get rid of the dependency?