Monocular object detector implementation for Waymo multi-view training and evaluation. Includes code for adaption, evaluation, data augmentation and configuration.
Hi ! This is really good repo and I am able to train the warmup on waymo.
However there is one issue I dont understand.
For depth aware , we need the prev_0 images right?
But the waymo Kitti adapter does not provide that?
How did you get it?
first of all, than you very much for this repo. I would be interested in the trained network and in the results that you achieved on the Waymo dataset. Can you please share it with me?