getCameraImagePrediction method

Future<String> getCameraImagePrediction(CameraImage cameraImage, { int? rotation, List<double> mean = torchVisionNormMeanRGB, List<double> std = torchVisionNormSTDRGB, CameraPreProcessingMethod cameraPreProcessingMethod = CameraPreProcessingMethod.imageLib, PreProcessingMethod preProcessingMethod = PreProcessingMethod.imageLib, })

Retrieves the top prediction label for a camera image.

Takes a cameraImage as input. Optional parameters include rotation, mean, std, cameraPreProcessingMethod, and preProcessingMethod. Returns a Future that resolves to a String representing the top prediction label.

Implementation

Future<String> getCameraImagePrediction(CameraImage cameraImage,
    {int? rotation,
    List<double> mean = torchVisionNormMeanRGB,
    List<double> std = torchVisionNormSTDRGB,
    CameraPreProcessingMethod cameraPreProcessingMethod =
        CameraPreProcessingMethod.imageLib,
    PreProcessingMethod preProcessingMethod =
        PreProcessingMethod.imageLib}) async {
  // Retrieve the prediction list for the camera image
  final List<double> prediction = await getCameraImagePredictionList(
      cameraImage,
      rotation: rotation,
      mean: mean,
      std: std,
      cameraPreProcessingMethod: cameraPreProcessingMethod,
      preProcessingMethod: preProcessingMethod);

  // Get the index of the maximum score from the prediction list
  int maxScoreIndex = softMax(prediction);
  // Return the label corresponding to the maximum score index
  return labels[maxScoreIndex];
}