I'd disagree, it probably got yours correct by partly random chance. Try testing a couple of selfies of yourself that are from a few days/weeks/minutes apart. This helps control the variables of facial structure and unique characteristics, and should also control for lighting and photo quality if you take pics consistently. When I tried that, it varied by 8 years. Unless my biological age rapidly goes up or down by years in the span of days, it's safe to conclude this tool is not reliable.
Ok, followed all the directions exactly and have re-checked all coding. All the installed files said they were successful installed. However, when program runs, I get a bunch of errors. one of which is, "ValueError: Face could not be detected in numpy array.Please confirm that the picture is a face photo or consider to set enforce_detection param to False." All four faces I put in the "faces" folder are valid faces in .jpg format.
Hey man i'm kind of curious on how you make your videos, do you know what you're typing out before you make the video and during the video you provide clarity and reasoning for the code? Or do you just raw-dog it like a straight G
help i allways get this kind of problem cv::findDecoder imread_('foto/ai.jpg'): can't open/read file: check file path/integrity Traceback (most recent call last): how to make deep face can read image from my folder?, i tried differnt ways to do it but result allways same, please help me
i have a problem i tried your first code but i had a problem raise JSONDecodeError("Expecting value", s, err.value) from None json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0) help please
Pretty sure DeepFace is awfully unreliable. I gave this a shot for fun and for an objective test I used the same person taken at relatively the same time. Selfie from myself separated by a few months varies by 8 years, two selfies taken within seconds of each other at relatively same angle and light varies by 6 years. For men it tended to guess an age lower than reality, but for women, about 10 years above over 70% of the time. Also seemed to be less certain if a girl was a girl or not, sometimes making surprising errors, whereas for men it was gave 99.9% certainty of the correct answer. FWIW, just tried it in default settings instead of playing with models. I'm curious what the reasons for this variation would be - is this likely due to insufficient training in the DeepFace model? My guess is it trained well with a set of images that were balanced in being more easy to distinguish than reality, and if given a huge sample size to test it probably does relatively well, yet still is strikingly far off from normal human classification ability. Though perhaps it does better on some certain face pictures that humans are poor at, IDK.
the algorithm tries to predict a gender, which has only 2 options, its not preference, its not pronounce, and its not feelings, or any other psychiatric disorder, its biological gender
I think that's the reason they removed microsoft's face age website. It seemed more accurate than this for age and sex, which for some people is unwelcome.
Hey 👋 I'm the creator of DeepFace. Thank you for sharing this to the community 🙏
great library.
Hello, where did you get the models ? As the ones I found are empty. Thanks in advance.
I checked out my age.
It is very accurate
Bro how did you tested it
Could you please guide me in the program execution
@@srivathsatantry8535 what is your problem?
I'd disagree, it probably got yours correct by partly random chance. Try testing a couple of selfies of yourself that are from a few days/weeks/minutes apart. This helps control the variables of facial structure and unique characteristics, and should also control for lighting and photo quality if you take pics consistently. When I tried that, it varied by 8 years. Unless my biological age rapidly goes up or down by years in the span of days, it's safe to conclude this tool is not reliable.
They just recently arrested a high school student for saying theres only 2 genders🤣🤣
Ok, followed all the directions exactly and have re-checked all coding. All the installed files said they were successful installed. However, when program runs, I get a bunch of errors. one of which is, "ValueError: Face could not be detected in numpy array.Please confirm that the picture is a face photo or consider to set enforce_detection param to False." All four faces I put in the "faces" folder are valid faces in .jpg format.
Hey man i'm kind of curious on how you make your videos, do you know what you're typing out before you make the video and during the video you provide clarity and reasoning for the code? Or do you just raw-dog it like a straight G
Great tutorial, learnt alot
help i allways get this kind of problem cv::findDecoder imread_('foto/ai.jpg'): can't open/read file: check file path/integrity
Traceback (most recent call last):
how to make deep face can read image from my folder?, i tried differnt ways to do it but result allways same, please help me
Awesome bro...Thanks 👍👍
Which software are u using
He, how can we see the probabilities of man or woman?
Does this identify genders like non-binary, two-sprited, or whatever gibberish kind too? 😂
What's two sprited lmaoo
😂😂😂
i have a problem
i tried your first code but i had a problem
raise JSONDecodeError("Expecting value", s, err.value) from None
json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)
help please
LOL
Vielleicht kannst du ein Video darüber machen, wie du deinen code planst? Von der Idee bis zum code. Grüß!
Thank you so much 😍
Sir after trying the steps you told in this video I'm getting 5 errors.
Does deepface require graphics card
Nice video. I tried to test the algorithm a bit by making it analyze people from Uzbekistan. That gave some rather hilarious results.
Love this content
So could you please tell me how to fix those errors. If you could plzz send me the source code then it would be very good sir please
Great tutorial. Seems those models works horrible on children photo
awesome channel!
DeepFace thinks Ariana Grande is a Man 🤣
Lol
Pretty sure DeepFace is awfully unreliable. I gave this a shot for fun and for an objective test I used the same person taken at relatively the same time. Selfie from myself separated by a few months varies by 8 years, two selfies taken within seconds of each other at relatively same angle and light varies by 6 years. For men it tended to guess an age lower than reality, but for women, about 10 years above over 70% of the time. Also seemed to be less certain if a girl was a girl or not, sometimes making surprising errors, whereas for men it was gave 99.9% certainty of the correct answer. FWIW, just tried it in default settings instead of playing with models. I'm curious what the reasons for this variation would be - is this likely due to insufficient training in the DeepFace model? My guess is it trained well with a set of images that were balanced in being more easy to distinguish than reality, and if given a huge sample size to test it probably does relatively well, yet still is strikingly far off from normal human classification ability. Though perhaps it does better on some certain face pictures that humans are poor at, IDK.
Thx_.
7th
This video will get reported for calling Emma a man instead of her preferred pronoun of "they/them".
the algorithm tries to predict a gender, which has only 2 options, its not preference, its not pronounce, and its not feelings, or any other psychiatric disorder, its biological gender
I think that's the reason they removed microsoft's face age website. It seemed more accurate than this for age and sex, which for some people is unwelcome.
Poor Emma being listed as a man 😝