search

bowenc0221 / panoptic-deeplab

This is Pytorch re-implementation of our CVPR 2020 paper "Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation" (https://arxiv.org/abs/1911.10194)
Apache License 2.0
590 stars 117 forks source link

Assertion Error when loading the cityscapes dataset #117

Closed ankur-chr closed 2 years ago

ankur-chr commented 2 years ago

I am getting the following error when the code tries to load the cityscapes dataset. Tried debugging this, but can't solve it. (running this on a conda based environment on a windows machine)

Traceback (most recent call last): File "tools/train_net.py", line 193, in main() File "tools/train_net.py", line 88, in main data_loader = build_train_loader_from_cfg(config) File "D:\Masters\dissertation\src\pdl\panoptic-deeplab\tools..\segmentation\data\build.py", line 113, in build_train_loader_from_cfg dataset = build_dataset_from_cfg(config, is_train=True) File "D:\Masters\dissertation\src\pdl\panoptic-deeplab\tools..\segmentation\data\build.py", line 83, in build_dataset_from_cfg dataset = dataset_map[config.DATASET.DATASET]( File "D:\Masters\dissertation\src\pdl\panoptic-deeplab\tools..\segmentation\data\datasets\cityscapes_panoptic.py", line 65, in init super(CityscapesPanoptic, self).init(root, split, is_train, crop_size, mirror, min_scale, max_scale, File "D:\Masters\dissertation\src\pdl\panoptic-deeplab\tools..\segmentation\data\datasets\cityscapes.py", line 84, in init assert len(self) == _CITYSCAPES_INFORMATION.splits_to_sizes[self.split] AssertionError

ankur-chr commented 2 years ago

Looks like this issue was due to missing leftImg8bit folder and content. Solved it!

ankur-chr commented 2 years ago

Solved