search

aimerykong / Low-Rank-Bilinear-Pooling

CVPR2017 - an ultra-compact bilinear model for fine-grained classification
https://www.cs.cmu.edu/~shuk/lr_bilinear.html
149 stars 36 forks source link

gender recognition based on the whole body #8

Open kclch opened 5 years ago

kclch commented 5 years ago

Can I use it for whole body based gender recognition?

aimerykong commented 5 years ago

Hi, if you have full body dataset, you can definitely train such models for that. It must be interesting to compare a simple feedforward model with the bilinear version.

On Wed, Sep 5, 2018 at 7:53 PM kclch notifications@github.com wrote:

Can I use it for whole body based gender recognition?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/aimerykong/Low-Rank-Bilinear-Pooling/issues/8, or mute the thread https://github.com/notifications/unsubscribe-auth/AGKZJKX8bGv7nDWITYbCzBilBTfELmbPks5uYI5EgaJpZM4WcBdX .

kclch commented 5 years ago

Thank you very much for your answer, but I will try to use the pedestrian image in all directions obtained from video surveillance as a data set. There is no guarantee that the face will be detected, and it will be divided into two categories, which are directly used by the source code. Is this method feasible?

aimerykong commented 5 years ago

It makes sense to do gender classification over whole body especially when the face is occluded or blurry. This should require you make the correct label (female&male) to enable good training. I would guess the model could learn to do it based on clothes/color. But it's interesting to see.

On Thu, Sep 6, 2018 at 6:20 AM kclch notifications@github.com wrote:

Thank you very much for your answer, but I will try to use the pedestrian image in all directions obtained from video surveillance as a data set. There is no guarantee that the face will be detected, and it will be divided into two categories, which are directly used by the source code. Is this method feasible?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/aimerykong/Low-Rank-Bilinear-Pooling/issues/8#issuecomment-419090572, or mute the thread https://github.com/notifications/unsubscribe-auth/AGKZJIADV3x0goo58haV8NA31hQ1HDYSks5uYSEtgaJpZM4WcBdX .

MrYanMYN commented 1 year ago

Is there any progress with creating such a network? I am currently looking for a solution and if there is an already trained alternative to self - making one from scratch that would be awesome!