Benchamrk on paper PixelFormer: Attention Attention Everywhere: Monocular Depth Prediction with Skip Attention
-
Download InSpaceType eval set. Install torch and torchivsion and packages: matplotlib, tqdm, tensorboardX, timm, mmcv, opencv-python
-
Download pretrained model 'nyu.pt' from Official Link and put it here
python pixelformer/test.py --data_path ./ --dataset nyu --filenames_file data_splits/split_files.txt --checkpoint_path nyu.pth --max_depth 10
The command generates report files for hierarchy (H0-H2). *-all means overall H0-H2 means level of hierarchy. H1_xx means scene space type number. See space_type_def.yml for space type number definition.