kitti gt interpolated for training?

isl-org / DPT

Dense Prediction Transformers

MIT License

1.96k stars 254 forks source link

kitti gt interpolated for training? #70

Open AntiLibrary5 opened 2 years ago

AntiLibrary5 commented 2 years ago

Hi, Do you finetune on kitti using the sparse GT as it is and interpolate predictions only for visualization or are the depth maps interpolated for training itself too (seemingly from the Resize transform method)?

awsaf49 commented 2 years ago

I am wondering the same thing.