Inference results

[LorenzoF6]

If i trained and tested my model, such as patchcore, on my own data and now i want to perform inference, is there any difference or is better one the other in making the inference with the engine.predict and using the openvino inferencer? And it is correct using, with the two different methods, the test data to perform the inference?
Thanks

[yxteo2]

You can check the torch inferencer in anomalib

If you check the code you can see actually you can use cv2 to load and pass the image to the model