Accuracy of model in TensorFlow Model Maker

How can I know the accuracy of my created model using the TensorFlow model maker?

for which task?

usually there’s a evaluate method that you run on your test data (more info here: Image classification with TensorFlow Lite Model Maker)

Object Detection. I also want to see my confusion matrix so I can see those positive predictions. is it possible?

I don’t know if the object detection task has a ready to use confusion matrix but you can copy the idea from the audio classification task: Transfer Learning for the Audio Domain with TensorFlow Lite Model Maker

Since the source code is open, you can, with some time, implement it on your side

thanks! I have another question. We created a mobile app for our object detection model, and I declare there, that if the confidence score is above 50% the object is acceptable. what are the factors to consider to support our claim saying 50% confidence is acceptable.