How to add custom normalize image data by MlContext.Transform ?
i have flow preprocess data in python like this : img -= (104, 117, 123) (RGB image) I want to do this with MlContext.Transform because i haved this : MLContext mlContext = new MLContext();
var pipeline = mlContext.Transforms.LoadImages(outputColumnName: "input0", imageFolder: "",
inputColumnName: nameof(ImageNetData.ImagePath))
.Append(mlContext.Transforms.ResizeImages(outputColumnName: "input0",
imageWidth: ImageNetSettings.imageWidth, imageHeight:
ImageNetSettings.imageHeight, inputColumnName: "input0"))
.Append(mlContext.Transforms.ExtractPixels(outputColumnName: "input0"))
.Append(mlContext.Transforms.ApplyOnnxModel(modelFile: modelLocation,
outputColumnNames: new[] { TinyYoloModelSettings.ModelOutputConf,
TinyYoloModelSettings.ModelOutputBox,
TinyYoloModelSettings.ModelOutputLandmark
}, inputColumnNames: new[] { TinyYoloModelSettings.ModelInput }));
How can i implement this custom normalize after Resize ???. Thank for your help.