CODEWITHSUNDEEP
pythonwindowstensorflowevalobject-detection-api
Luiggi
Luiggi

Reputation: 11

how to find the model precision Faster_rcnn_inception_v2?

please help .

run eval.py of the tensorflow detection model

I want to find the precision and I got this data. Can someone explain to me if it's ok or not and what can I do please.

I am new in these subjects

enter image description here

Upvotes: 1

Views: 292

Answers (1)

user11530462
user11530462

Reputation:

To explain the concepts in detail.

  1. For object detection instead of precision Average precision is considered.
  2. AP (Average Precision) is a popular metric in measuring the accuracy of object detectors.
  3. Average precision computes the average precision value for recall value over 0 to 1.

So what you need to do in your case is to create a DataFrame with Objects, Actual value and Prediction value, the prediction you can say TRUE if your IoU value >=0.5.

IoU measures the overlap between 2 boundaries. We use that to measure how much our predicted boundary overlaps with the ground truth (the real object boundary). In some datasets, we predefine an IoU threshold (say 0.5) in classifying whether the prediction is a true positive or a false positive.

enter image description here

Then you can calculate the average precision accordingly.

Hope this answer helps you, Happy Learning!

Upvotes: 1

Related Questions