MLHandActionClassifier.DataSource.labeledFiles(at:)
Creates a data source from a folder that contains videos, each named after the hand action they represent.
Declaration
case labeledFiles(at: URL)Parameters
- at:
The URL to a folder in the file system that contains folders of hand action videos. The data source uses each file’s name before the first period as its classification label.
Discussion
Create a hand action data source from a directory of videos with the labeledFiles case. You must name each video file with a hand action label, followed by a period and an arbitrary string, ending with the video file’s extension. For example, you can name a hand action classifier’s training files Wave.3.mov, MoveCloser.1.mov, MoveCloser.2.mov, and so on.
In this example, a hand action classifier would have at least two class labels:
WaveMoveCloser
See Also
Creating a data source
MLHandActionClassifier.DataSource.labeledDirectories(at:)MLHandActionClassifier.DataSource.directoryWithVideosAndAnnotation(at:annotationFile:videoColumn:labelColumn:startTimeColumn:endTimeColumn:)MLHandActionClassifier.DataSource.labeledVideoDataFrame(_:videoColumn:labelColumn:startTimeColumn:endTimeColumn:)MLHandActionClassifier.DataSource.labeledVideoData(table:videoColumn:labelColumn:startTimeColumn:endTimeColumn:)MLHandActionClassifier.DataSource.labeledKeypointsDataFrame(_:sessionIdColumn:labelColumn:featureColumn:)MLHandActionClassifier.DataSource.labeledKeypointsData(table:sessionIdColumn:labelColumn:featureColumn:)