otx.core.data.adapter#

OTX Core Data Adapter.

Functions

get_dataset_adapter(task_type, train_type[, ...])

Returns a dataset class by task type.
otx.core.data.adapter.get_dataset_adapter(task_type: TaskType, train_type: TrainType, train_data_roots: str | None = None, train_ann_files: str | None = None, val_data_roots: str | None = None, val_ann_files: str | None = None, test_data_roots: str | None = None, test_ann_files: str | None = None, unlabeled_data_roots: str | None = None, unlabeled_file_list: str | None = None, **kwargs)[source]#

Returns a dataset class by task type.

Parameters:

  • task_type – A task type such as ANOMALY_CLASSIFICATION, ANOMALY_DETECTION, ANOMALY_SEGMENTATION, CLASSIFICATION, INSTANCE_SEGMENTATION, DETECTION, CLASSIFICATION, ROTATED_DETECTION, SEGMENTATION.

  • train_type – train type such as Incremental and Selfsupervised. Selfsupervised is only supported for SEGMENTATION.

  • train_data_roots – the path of data root for training data

  • train_ann_files – the path of annotation file for training data

  • val_data_roots – the path of data root for validation data

  • val_ann_files – the path of annotation file for validation data

  • test_data_roots – the path of data root for test data

  • test_ann_files – the path of annotation file for test data

  • unlabeled_data_roots – the path of data root for unlabeled data

  • unlabeled_file_list – the path of unlabeled file list

  • kwargs – optional kwargs

Modules

otx.core.data.adapter.action_dataset_adapter

Action Base / Classification / Detection Dataset Adapter.

otx.core.data.adapter.anomaly_dataset_adapter

Anomaly Classification / Detection / Segmentation Dataset Adapter.

otx.core.data.adapter.base_dataset_adapter

Base Class for Dataset Adapter.

otx.core.data.adapter.classification_dataset_adapter

Classification Dataset Adapter.

otx.core.data.adapter.detection_dataset_adapter

Detection Dataset Adapter.

otx.core.data.adapter.segmentation_dataset_adapter

Segmentation Dataset Adapter.

otx.core.data.adapter.visual_prompting_dataset_adapter

Visual Prompting Dataset Adapter.