Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

how to get the precision and recall from the Evaluation Script URGENT !!! #9

Open
Hiteshsaai opened this issue Aug 31, 2020 · 6 comments

Comments

@Hiteshsaai
Copy link

I want to know which part of the code is calculating the precision and recall in the evaluation script.

is this line of code gives the precision and recall, and i especially want the recall values, the below code gives values for 1000 points. does that mean it is for 1000 images ?? URGENT !!

propose = pr_curve[:, 0]
recall = pr_curve[:, 1]
@wondervictor
Copy link
Owner

@Hiteshsaai 1000 points for evaluation, not 1000 images.

@Hiteshsaai
Copy link
Author

@wondervictor what do you mean by points and can you tell me whether this line of code below is the right one to get the precision and recall, if not can you help me with getting them from the evaluation script?

propose = pr_curve[:, 0]
recall = pr_curve[:, 1]

@SURABHI-GUPTA
Copy link

@wondervictor @Hiteshsaai I want to calculate detection accuracy. Can you explain the logic for calculating precision and recall in your code ?

@Hiteshsaai
Copy link
Author

@SURABHI-GUPTA you can get the recall & precision from the evaluation function of the evaluation.py script.

Declare a new array for precision and recall and Append propose and recall values for every iteration into the respective array in the evaluation function at the end after line number 260, and return those two arrays.

Each array will have three nested array inside it like ---> [[],[],[]]
using that you can calculate precision and recall.

@watertianyi
Copy link

@wondervictor I don’t understand why I set 1000 points for calculation, and I don’t know how to calculate the reasonable rate and value rate. What is the principle of this design?

@gkrisp98
Copy link

Has anyone figured it out ?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants