object detection algorithm outputs an image with no bounding boxes around the detected objects after testing the algorithm

Question

my partner and I are have created a modified faster rcnn algorithm and it doesn't seem to have any bounding boxes when it outputs an image after testing the algorithm but it only outputs the Mean Average Precision. so far what I've tried doing is uninstalling and reinstalling TensorFlow but the problem still persists.

Answer 1

Try changing the confidence Threshold so, that you are able to see bounding boxes having lower confidences, this way you can verify whether there is a problem with the postprocessing code or your model is not able to learn any useful patterns.

Also please provide some insight to your code, so that we can help better