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 can i recognize faces instead of canvas instead of simple buffer in node? #174

Closed
p3x-robot opened this Issue Dec 27, 2018 · 7 comments

Comments

Projects
None yet
2 participants
@p3x-robot
Copy link

p3x-robot commented Dec 27, 2018

hello,

how are you?
i am trying to create a home made recognition system, that can open the door , turn on the lamps etc ...
but if i am on raspberry pi , i probably wont be able to use the canvas npm and it has a lot overhead. is there any way to give the image input a simple buffer?
thanks for your response,
patrik

@justadudewhohacks

This comment has been minimized.

Copy link
Owner

justadudewhohacks commented Dec 27, 2018

Hi,

Sure you can pass a tensor instead of a canvas as input to the neural nets as well.

@p3x-robot

This comment has been minimized.

Copy link

p3x-robot commented Dec 27, 2018

so you think if i create a 3d matrix is enough - RGB?

@justadudewhohacks

This comment has been minimized.

Copy link
Owner

justadudewhohacks commented Dec 27, 2018

Yes, the nets expect RGB input, so you can create either tensors of shape [height, width, channels] or [batchSize, height, width, channels] for batch input. Note batch inputs don't work for the face detection nets currently.

@p3x-robot

This comment has been minimized.

Copy link

p3x-robot commented Dec 27, 2018

ciao,
so do you think if i only want to use on server i should rather use opencv4nodejs or face-api.js is still better to use?

@justadudewhohacks

This comment has been minimized.

Copy link
Owner

justadudewhohacks commented Dec 28, 2018

opencv4nodejs is for general image processing. If you only need face detection, recognition, classification etc. use face-api.js

@p3x-robot

This comment has been minimized.

Copy link

p3x-robot commented Dec 28, 2018

ok, and another question, i do not have to train at all? one image and the face distance is enough, right?

@justadudewhohacks

This comment has been minimized.

Copy link
Owner

justadudewhohacks commented Dec 28, 2018

Exactly.

@p3x-robot p3x-robot closed this Dec 28, 2018

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