Skip to content
This repository was archived by the owner on Jun 18, 2024. It is now read-only.

Conversation

@AshishMaru
Copy link

Unknown faces temporary detection added by training one more label that includes 7 Images of all training data, I tested this with more than 50 faces and it's working fine

Unknown faces temporary  detection added by training one more label that includes 7 Images of all training data, I tested this with more than 50 faces and it's working fine
@Aman2612
Copy link

Tested this but it still classifies the unregistered users as someone already registered.

@AshishMaru
Copy link
Author

I used tensor flow with SVM and it's working very well for me and also if I capture images in good light then 100% time it is working well with the unregistered person.

@Aman2612
Copy link

I see that you modifed only the training file. Did you modify any setting or let it remain default? anything else?

@AshishMaru
Copy link
Author

No, I've not modified any other setting and if you need better result then please train at least 5 faces(20 images/ face).
and use at least 15 images for unknown faces,
changes the
if (i <7) {
arrayUnknownFace.add(file);
}

to

if (i <15) {
arrayUnknownFace.add(file);
}

In TrainingActivity.java

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants