Face Recognition in Python | Part 1 | FaceNet, MTCNN, SVM

[@fdossantos98]

I paid a couse on Udemy about the same subject, your video is much better. Thank you!

[@OJRO-xp2ip]

You are a life saver bro. Very accurate and easily understandable.

[@sardorabdirayimov]

Glad it helped

[@hatake_kakashi165]

i was making a model for my college for attendance this content helped a lot thank you

[@harshids]

Please help how to do iam also makeing a model for my college

[@ElenaEroshina]

This is the most useful tutorial I've seen befor. Thank you so much for it. It really helped ☺

[@sardorabdirayimov]

Happy to hear that!

[@john94139]

Thank you. Your tutorial was very useful for me!

[@3agoskin]

Thank you for this video! Very helpful!

[@dexterjeong]

Thank you for the great lecture. 👍

[@djamaltoe3774]

I will speak in french... Tu es fort

[@djamalyoussouftoe]

What happens if the face I want to predict is note in my faces dir ? I mean that if I have Robert, Anna , and Prince. Now I put Hercules image, normale I must put Unknown no ? Must I calculate accuracy ? And avoid prédictions under .90 ?

[@khumoyunmirzosirojov9998]

Keep going 👍👍👍

[@sardorabdirayimov]

Thank you, I will

[@qorganbek_d]

Salam from Kazakhstan bro

[@samandarkhodjiyev5658]

good job

[@koyiljonvaliyev4076]

Mashalloh

[@Ankushkumar-uk9wj]

i am stuck.. please help? Q. how to lable an image an 'UNKNOWN' if the image is not in the dataset...? as model tries to predict the unknown and random image also? please reply

[@primitarahmanifertiasimabu1159]

i have question sir.. how to make this file to extension .py? because i want to implementation this to my app for training section using flask..I'm confused by the order of the code

[@Amazondealsin-rq7do]

model = SVC(kernel='linear', probability=True) model.fit(X_train, Y_train) ValueError: The number of classes has to be greater than one; got 1 class Sir could you please explain ,its important

[@sardorabdirayimov]

It seems that Y_train contains only 1 class values. Recheck the Y_train values

[@nagarajuchakrapani995]

Can it identify multiple persons in a frame/photo

[@hidayatullahzaheen7761]

i make a face recognition attendance system in python that work on face to present students but here is a problem that store students many times until the face is detecting how can write a function that store a student just ones in 24 hours

[@sardorabdirayimov]

Hi there, Sorry for late response. I have no experience with developing such function. kzbin.info/aero/PLRqwX-V7Uu6aG2RJHErXKSWFDXU4qo_ro&si=SAjBwfFmFBL_UnQw I hope the playlist might be useful Happy learning

[@imenmabrouk6769]

what if there is a total unkown and we want to classify that person is unkown?

[@sardorabdirayimov]

Here is my comment on @imenmabrouk6769: My SVM is trying to classify the input face into the trained faces. If you want to add unknown, I suggest you to use “softmax” activation function which return the probability of given face to the trained face. If you give mode the new face, it will return like taylor swift: 0.3 , person B: 0.5. Then you can create conditions that if probability is lower than 0.7 (as example) then it is “unknown”

[@jlkun5918]

is MTCNN is a pre-trained dataset?

[@MrKhashaz]

Thanks for sharing. Appreciate it. Can you help and share How to do face embedding from aligned image of faces? I have image that have been cropped. Can you share how to get the x,y,w,h from all frame image not from mtcnn detect face. Thanks for help anyway.

[@sardorabdirayimov]

Try opencv python image.shape to get height width and channel. While in order to find x,y you need to find it manually or use model. If you want full image then set x and y as 0

[@MrKhashaz]

@@sardorabdirayimov thanks for reply and help. sorry if I'm still newbie and still learn on this. Could you please to share example of code on this: lets say I have image-cropped.jpg with size 120x120, I want to set all image as input. at your function to extract_face how to load full image as the input. ---- def extract_face(self, filename): img = cv.imread(filename) img = cv.cvtColor(img, cv.COLOR_BGR2RGB) x,y,w,h = self.detector.detect_faces(img)[0]['box'] x,y = abs(x), abs(y) face = img[y:y+h, x:x+w] return face_arr ---- I'm confusing what should x,y.w,h if for full image size? and how about x,y? Appreciate your help.

[@user-bm1ti6ok6h]

i am facing one error this line X,Y = faceloading.load_classes(), i can not understand, please help me sir. it's my humble request.i am waiting for your reply. it is very important sir. please help me sir showing this error --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) in () 1 faceloading = FACELOADING("/content/Dataset") ----> 2 X,Y = faceloading.load_classes() AttributeError: 'FACELOADING' object has no attribute 'load_classes'

[@arunabhadutta72]

I need serious help in this regard as I am getting an error in test_im = [test_im] ypreds = model.predict(test_im) value error: found array with dim 3. SVC expected

[@sardorabdirayimov]

Do not increase the dimension by adding [ ]. According to your code, it expand existing spaces Hope it helps

[@Amazondealsin-rq7do]

could you please explain it clearly , sir @@sardorabdirayimov

[@sardorabdirayimov]

@@Amazondealsin-rq7do Initially test_im array has 2 dimensions. test_im=[test_im] >> this code makes dimension equal to 3. Which is an issue that SVM cannot solve

[@eixadarko5380]

thanks dude you help me so much can i talk to you on any social app like face book iwant to ask some question and i have graduation project and thank you duude

[@sardorabdirayimov]

Happy to be helpful. You can contact me at email [e-mail]

[@nimshashajahan4357]

Is this code also be implemented using live video

[@sardorabdirayimov]

For sure))

[@azaman_s]

ozbek tilida shiqaring oka