The MXNet Scala Infer API provides you with model loading and inference functionality using the MXNet Scala package.
To use the Infer API you must first install the MXNet Scala package. Instructions for this are provided in the following variations:
The Scala Infer API includes both single image and batch modes. Here is an example of running inference on a single image by using the ImageClassifier class. A complete image classification example using ResNet-152 is provided in the Scala package's example folder. This example also demonstrates inference with batches of images.
def runInferenceOnSingleImage(modelPathPrefix: String, inputImagePath: String, context: Array[Context]): IndexedSeq[IndexedSeq[(String, Float)]] = { val dType = DType.Float32 val inputShape = Shape(1, 3, 224, 224) val inputDescriptor = IndexedSeq(DataDesc("data", inputShape, dType, "NCHW")) // Create object of ImageClassifier class val imgClassifier: ImageClassifier = new ImageClassifier(modelPathPrefix, inputDescriptor, context) // Loading single image from file and getting BufferedImage val img = ImageClassifier.loadImageFromFile(inputImagePath) // Running inference on single image val output = imgClassifier.classifyImage(img, Some(5)) output }