Error while Running with KITTI dataset.

[terrybelinda]

Hello, Thank you for your amazing work. We are students trying to inference this against KITTI dataset. However, we keeping running into the following issue when we running the following command.

!python test.py --gpu 0 --workdir /content/drive/MyDrive/ShapeNetResults9 --model grnet --weights /content/drive/MyDrive/GRNet-KITTI.pth --test_mode kitti

Error: File "test.py", line 80, in main() File "test.py", line 76, in main model.test() File "/content/SpareNet/runners/base_runner.py", line 351, in test self.val() File "/content/SpareNet/runners/base_runner.py", line 209, in val self.val_step(items) File "/content/SpareNet/runners/grnet_runner.py", line 59, in val_step _, _, refine_ptcloud, coarse_loss, refine_loss = self.completion(data) File "/content/SpareNet/runners/grnet_runner.py", line 80, in completion coarse_loss = self.chamfer_dist_mean(coarse_ptcloud, data["gtcloud"]).mean() KeyError: 'gtcloud'

We noticed, that data has only "partial box" and "bounding box" and not ground truth. Is there anything we must change ?

[AlphaPav]

Thanks for reporting this issue! During testing, actually loss is not required. You can delete the lines related to the loss and run again. Note that we trained a model on the ShapetNet Car dataset and then use this pre-trained model to do inference on the KITTI dataset.