Demo of using Unity Sentis (Open Beta) to perform inference for image classification task

A pretrained resnet model was converted to the ONNX format using Pytorch's internal ONNX converter. This model performs image classification. In the demo we test it by inferencing on a cat and a dog image.

General steps to run inference on an ML model

• import the ONNX file into Unity
• create a runtime model (ie, representation of the model in Unity Sentis)
• create an engine to run the model
• run the model to get output
• get and print output from the model

Inference Result

[]