Asking about deployment of onnx format models

[Kyle-fang]

Please ask! What is the final output obtained by converting the trained model into onnx format and importing it into onxxruntime for inference? Is it the input sample detection result (and whether it is an anomalous sample) and the corresponding confidence level? Or is it the anomaly_map with the same channel 1 as the input height and width?

[Apostatee]

you should rewrite an ensemble model which put the three models together, and the model output the anomaly score map directly.