How to check whether my trained model detects object correctly by coding?

Question

I am using my trained model to test images. The model can detect objects of my test images in the right way, but not all of them, so I would like to ask how can I check if it detects the right objects in an image. It 's so easy to do that by using my eyes but I want to do it by coding. Is there any options to do that? Thanks.

I am using Tensorflow Object detection API and Python 3.6.

Answer 1

You are talking about benchmark metrics, usually, mAP (mean average precision) is used. Check mAP metric in object detection and computer vision