Ask Your Question

FaceRecognizer: how to train

asked 2012-09-05 08:50:15 -0500

updated 2012-09-05 09:39:38 -0500

Andrey Pavlenko gravatar image

i am using ubuntu system. I tried to use the sample python script( for facerecogniton. However i have a error that i have to train at least a images. How do i train images? I have created a csv file that contain the path to the dataset that the tutorial teaches to use.

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2012-09-05 12:09:57 -0500

For the Python tutorials on cv::FaceRecognizer see the following posts:

I've tried to explain it in great detail, but please don't hesitate to ask if something is unclear. These tutorials will be added to the OpenCV tutorials, as soon as I find some time. You don't need to create a CSV File for these tutorials.

edit flag offensive delete link more


i have successfully train the images. But how do i match a images that i have with the trained images and give a score for its accuracy?

hittourou gravatar imagehittourou ( 2012-09-11 20:10:50 -0500 )edit

What's wrong with the FaceRecognizer.predict as shown in all the demos and documentation? For example

Philipp Wagner gravatar imagePhilipp Wagner ( 2012-09-12 01:13:41 -0500 )edit

Whenever i try my own images it will have this error RuntimeWarning: invalid value encountered in divide X = X / float((maxX - minX)) and it will always be Predicted label = 0 (confidence=0.00) I still do not understand how this facerecognizer works. I thought i have to train images to the data and store it there, if a input images match with the data then the then the predicted label will be 0?

hittourou gravatar imagehittourou ( 2012-09-12 19:57:22 -0500 )edit
Login/Signup to Answer

Question Tools


Asked: 2012-09-05 08:50:15 -0500

Seen: 1,021 times

Last updated: Sep 05 '12