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

I think this ObjectDetection2DSet Comment May be Incorrect #324

Open
derkmed opened this issue Oct 26, 2023 · 4 comments
Open

I think this ObjectDetection2DSet Comment May be Incorrect #324

derkmed opened this issue Oct 26, 2023 · 4 comments

Comments

@derkmed
Copy link
Contributor

derkmed commented Oct 26, 2023

# [num_detections, len(label_vec)]

When subscribing to topics that publish this message, I did not read a flattened matrix. Instead I read a vector of confidence scores corresponding to the label_vec.

@derkmed derkmed changed the title ObjectDetection2DSet Comment May be Incorrect I think this ObjectDetection2DSet Comment May be Incorrect Oct 26, 2023
@Purg
Copy link
Member

Purg commented Oct 26, 2023

This comment is "correct", but a previous incarnation of the object detector was populating the message incorrectly. The newest Yolo detector that we have added should now be populating this field correctly.

@Purg
Copy link
Member

Purg commented Nov 17, 2023

@derkmed Checking in if this explanation was sufficient to close this issue or if there is still something to be resolved.

@derkmed
Copy link
Contributor Author

derkmed commented Dec 4, 2023

Got it. I will double check @Purg . So, each row should correspond to a detected object indicated via one-hot vector. Is that correct?

@Purg
Copy link
Member

Purg commented Dec 5, 2023

Got it. I will double check @Purg . So, each row should correspond to a detected object indicated via one-hot vector. Is that correct?

Yes, but the vector is not specifically constrained to be one-hot: it is a per-class confidence vector. Some detector implementations may choose to populate this vector in a one-hot fashion.

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

2 participants