vis4d.zoo.base.datasets.bdd100k.sem_seg

BDD100K dataset config for semantic segmentation.

Functions

get_bdd100k_sem_seg_cfg([data_root, ...])	Get the default config for BDD100K semantic segmentation.
get_test_dataloader(data_root, anno_path[, ...])	Get the default test dataloader for BDD100K segmentation.
get_train_dataloader(data_root, anno_path[, ...])	Get the default train dataloader for BDD100K segmentation.

get_bdd100k_sem_seg_cfg(data_root='data/bdd100k/images/10k', train_split='train', train_keys_to_load=('images', 'seg_masks'), test_split='val', test_keys_to_load=('images', 'seg_masks'), data_backend=None, image_size=(720, 1280), crop_size=(512, 1024), samples_per_gpu=2, workers_per_gpu=2)

Get the default config for BDD100K semantic segmentation.

Return type:

DataConfig

get_test_dataloader(data_root, anno_path, keys_to_load=('images', 'seg_masks'), data_backend=None, image_size=(720, 1280), samples_per_gpu=1, workers_per_gpu=1)

Get the default test dataloader for BDD100K segmentation.

Return type:

ConfigDict

get_train_dataloader(data_root, anno_path, keys_to_load=('images', 'seg_masks'), data_backend=None, image_size=(720, 1280), crop_size=(512, 1024), samples_per_gpu=2, workers_per_gpu=2)

Get the default train dataloader for BDD100K segmentation.

Return type:

ConfigDict