GilLevi / AgeGenderDeepLearning

628 stars 284 forks source link

Please help me to improve performance #20

Closed openedhardware closed 6 years ago

openedhardware commented 6 years ago

@GilLevi I am detecting gender/age from the video, and some faces aren't working well. Here are sample images(you could see the age at the end of file name)

male_0_2 Male, (0, 2)

1506005830019_hkc213254dfead43_0 06_male_ 0 2

(Male, (0, 2))

1506005307778_hkc213254dfead43_23 41_male_ 0 2

(Male, (0, 2))

1506006046966_hkc213254dfeaab1_17 79_male_ 4 6

(Male, (4, 6))

1506006314181_hkc213254dfeaab1_1 92_female_ 0 2

(Female, (0, 2))

Any suggestion to improve performance?

Do I crop images too large? Do I need to rotate image before getting age?

Thanks!

GilLevi commented 6 years ago

Hi @openedhardware,

Thank you for interest in our model.

I think the crop is indeed to large and needs to be more aggressive.

Best, Gil

openedhardware commented 6 years ago

Yes, @GilLevi That fixed the issue! Thanks!