hello friend

You can used methods based on iris detection and dynamic image processing techniques...

I hope this helps you.

Best regards

Dr.Indrajit Mandal

Dear Indrajit

Thanks for your answer.

In your opinion, what's about based on skin color methods?

Viola–Jones object detection framework is suitable for face detection. But gender detection needs some more enhancement in this framwokk

Dear Mahmoud and Sundaram,

Thanks for your answer

But other method is not for realtime face detection.

For real time face detection , you can do it with the help of OPENCV and a kit.

I've used Viola-Jones as implemented in OpenCV successfully

What would You think about "Inverse compositional Active Appearance Models methods". (ic-AAM). I'm not very expert about but it seems to be very up-and-coming.

Of course it would be very affected from lighting (like all appearance based methods) but if we could find a lighting normalization...

By the way I think it depends on application. Usually the best solution is also an Hybrid solution.

http://www.mathworks.com/matlabcentral/fileexchange/32704-icaam-inverse-compositional-active-appearance-models.

https://www.youtube.com/watch?v=_ssr4Ds9Aj0&list=UUSrHYe81P6cQcagP32C1cjA

Dear Leonardo,

Thank you.

Lighting is one of the main challenges in face detection.

You could try the following software library and contact developers for models specification:

http://www.iis.fraunhofer.de/content/dam/iis/de/documents-live/Forschungsfelder/Bildsysteme/Bildsensorik/SHORE_Flyer.pdf

http://www.iis.fraunhofer.de/de/ff/bsy/tech/bildanalyse/shore-gesichtsdetektion.html