Exportable Code

Exportable code for Anomaly tasks.

class otx.algorithms.anomaly.adapters.anomalib.exportable_code.AnomalyBase(model_adapter, configuration=None, preload=False)

Wrapper for anomaly tasks.

classmethod parameters()

Dictionary containing model parameters.

class otx.algorithms.anomaly.adapters.anomalib.exportable_code.AnomalyClassification(model_adapter, configuration=None, preload=False)

Wrapper for anomaly classification task.

postprocess(outputs: Dict[str, ndarray], meta: Dict[str, Any]) → float

Resize the outputs of the model to original image size.

Args:

outputs (Dict[str, np.ndarray]): Raw outputs of the model after infer_sync is called. meta (Dict[str, Any]): Metadata which contains values such as threshold, original image size.

Returns:

float: Normalized anomaly score

class otx.algorithms.anomaly.adapters.anomalib.exportable_code.AnomalyDetection(model_adapter, configuration=None, preload=False)

Wrapper for anomaly detection task.

postprocess(outputs: Dict[str, ndarray], meta: Dict[str, Any]) → ndarray

Resize the outputs of the model to original image size.

Args:

outputs (Dict[str, np.ndarray]): Raw outputs of the model after infer_sync is called. meta (Dict[str, Any]): Metadata which contains values such as threshold, original image size.

Returns:

np.ndarray: Detection Mask

class otx.algorithms.anomaly.adapters.anomalib.exportable_code.AnomalySegmentation(model_adapter, configuration=None, preload=False)

Wrapper for anomaly segmentation task.

postprocess(outputs: Dict[str, ndarray], meta: Dict[str, Any]) → ndarray

Resize the outputs of the model to original image size.

Args:

outputs (Dict[str, np.ndarray]): Raw outputs of the model after infer_sync is called. meta (Dict[str, Any]): Metadata which contains values such as threshold, original image size.

Returns:

np.ndarray: Segmentation Mask