Contents

MLActionClassifier.DataSource.labeledKeypointsDataFrame(_:sessionIdColumn:labelColumn:featureColumn:)

A data source made up of keypoints in a data frame.

Declaration

case labeledKeypointsDataFrame(DataFrame, sessionIdColumn: String = __Defaults.sessionIdColumnName, labelColumn: String = __Defaults.labelColumnName, featureColumn: String = __Defaults.featureColumnName)

Parameters

  • dataFrame:

    A DataFrame containing keypoints and labels.

  • sessionIdColumn:

    The name of the column containing session identifiers. Defaults to “session_id”.

  • labelColumn:

    The name of the column containing the labels. Defaults to “label”.

  • featureColumn:

    The name of the column containing the keypoints. Defaults to “keypoints”.

Discussion

The data frame must contain a column of session identifiers, a column of labels, and a column of keypoints. Each set of keypoints must be a multi-dimensional 1x3x18 array that contains the x, y, and z coordinates of each of the 18 keypoints. See VNRecognizedPointsObservation for more details.

See Also

Extracting key points