Comment on page
In Three Simple Steps
Computer vision is perhaps the most common use case for machine learning. ML allows for solving some interesting vision tasks, like image classification, detection, segmentation, and so on. Across all of these tasks, the NatML workflow is relatively uniform:
First, we create a model:
// Fetch the SSD Lite object detection model
var model = await MLEdgeModel.Create("@natsuite/ssd-lite");
We can log the model to see what data it consumes and produces:
// Inspect the model
// Input: input.1: (1, 3, 300, 300) System.Single <-- INPUT IMAGE
// Output: 836: (1, 3000, 21) System.Single
// Output: 863: (1, 3000, 4) System.Single
Then we instantiate the predictor class for the model:
// Create the SSD Lite predictor
var predictor = new SSDLitePredictor(model);
If you are using your own model file, you will have to write a predictor class implementation for it.
You will always want to create the model and predictor once, usually in your
Startmethod, because it is a time-consuming process.
To make a prediction, we simply need to call
Predicton the predictor:
// Given an image...
Texture2D image = ...;
// Make a prediction
var results = predictor.Predict(imageFeature);
Some predictors might require multiple inputs. See the
READMEof your predictor for what features to pass in.
Some predictors require explicit disposal when they are no longer needed. These predictors define a
// Dispose the predictor
predictor.Dispose(); // this may or may not exist, so refer to your predictor README
Once the predictor is disposed, the model MUST be disposed.
// Dispose the model