Ask Your Question
0

gender detection, simple question

asked 2016-05-16 05:08:13 -0500

atv gravatar image

I understand the tutorials, but i don't quite understand how to make it work for me. I could make a simple program, put names in a database, and give them a label 1 or 2, for male or female.

  1. Should i put in my own recognized/trained people, or use a large dataset.
  2. If the first, it feels a bit like cheating, as i am basically labeling that person with a gender already. I'm sure it will tell me i am a male if i labeled it as such and there are enough images of me.

I guess i dont understand the difference between just entering a hardcoded entry of "male" or "female" and actually do prediction.

Hope this makes sense :-)

atv

edit retag flag offensive close merge delete

Comments

The cheating can be ignored, because you should always use different training and validation data and thus, this means you will never provide the training label for a testing sample.

StevenPuttemans gravatar imageStevenPuttemans ( 2016-05-17 08:14:36 -0500 )edit
1

Hey Steven, thanks for your comment. So you mean i should always use a outside database like Yale or similar for facial expression, and never my own csv (of people i have trained to recognize).

That would make sense.

atv gravatar imageatv ( 2016-05-17 08:19:05 -0500 )edit

Yep because you are doing it for genders and thus want a large set to diverse over individuals.

StevenPuttemans gravatar imageStevenPuttemans ( 2016-05-17 09:18:19 -0500 )edit
1

Gotcha, thanks Steven

atv gravatar imageatv ( 2016-05-17 09:32:45 -0500 )edit

1 answer

Sort by ยป oldest newest most voted
2

answered 2016-05-16 10:17:32 -0500

IMHO, the tutorial's goal is to provide you a sample that can work. It is not a solution for any real life application. However, to make it achieve as higher result as possible, you should:

  1. Use training images that have the same conditions (light, outdoor conditions, etc.) as the test images your application will manipulate.
  2. It is not cheating. To recognize any kind of object (in this case, gender), your application must base on some model which is derived from correct training data and this mechanism is similar to our brain. This is actual prediction.
edit flag offensive delete link more
Login/Signup to Answer

Question Tools

1 follower

Stats

Asked: 2016-05-16 05:08:13 -0500

Seen: 131 times

Last updated: May 16 '16