getImagePredictionListFromBytesList method

Future<List<double>> getImagePredictionListFromBytesList(
  1. List<Uint8List> imageAsBytesList,
  2. int imageWidth,
  3. int imageHeight, {
  4. List<double> mean = torchVisionNormMeanRGB,
  5. List<double> std = torchVisionNormSTDRGB,
})

Returns a list of predictions for an image as a bytes list. The image are passed as a list of Uint8List objects. The imageWidth and imageHeight parameters specify the dimensions of the image. The optional mean and std parameters can be used to normalize the image. Returns a Future that resolves to a list of double values representing the predictions.

Implementation

Future<List<double>> getImagePredictionListFromBytesList(
    List<Uint8List> imageAsBytesList, int imageWidth, int imageHeight,
    {List<double> mean = torchVisionNormMeanRGB,
    List<double> std = torchVisionNormSTDRGB}) async {
  // Assert mean std
  assert(mean.length == 3, "Mean should have size of 3");
  assert(std.length == 3, "STD should have size of 3");

  // Call the getImagePredictionList method of the ModelApi class to get the predictions
  final List<double> prediction = (await ModelApi().getImagePredictionList(
          _index, null, imageAsBytesList, imageWidth, imageHeight, mean, std))
      .whereNotNull()
      .toList();

  return prediction;
}