Face Recognition in Python using face_recognition Library (in Google Colab)

We use the popular face_recognition library to recognize faces in new images. I explain about how it works briefly and write implement using Google Colaboratory, which provides GPU support. This video is for beginners who are looking to get started with face recognition. You can find many more examples in the official github repository of face_recognition.
GitHub Repository: github.com/age...
Explanation post by Adam Geitgey: / machine-learning-is-fu...
Subscribe to my Newsletter !
adarsh1021.git...
Having trouble ? Need help ? Connect with me !
Email: [e-mail]
Twitter : / adarsh_menon_
LinkedIn: / adarsh-me. .
Github : github.com/ada...
#face_recogniton #opencv #python

[@AdarshMenon]

Here is the original Colab - colab.research.google.com/drive/15_yQKwhVsmfN2p7geco_VFGKLYqBah04?usp=sharing

[@siddhipandare3882]

This is the best video for face recognition! Very well explained :)

[@thenotsomebody]

this is the best face recognition video

[@anefuoche1053]

i get an IndexError: list index out of range, that points to the for loop for known_dir; img_enc = face_recognition.face_encodings(img)[0]... but it works fine when i omit the index [0]. also in the for loop for unknown_dir i get a ValueError: operands could not be broadcast together with shapes (41,) (128,) that points to results = face_recognition.compare_faces(known_encodings, img_enc)...please what should I do?

[@iputubastianadiputra5816]

great explination ma brother

[@jaisonjose8256]

from my experience, for personal use. Training the data using keras by our images is more efficient

[@AdarshMenon]

I agree, this was more of a getting started type of video because many people were requesting for face rec after face detection

[@jaykaran2278]

It becomes tedious to train for every new face the system detect. Don't you think?

[@jaisonjose8256]

I specified 'personal'

[@chittiprolusunilkumar3975]

good work and it help for learn to kick start thank you and well explained

[@AdarshMenon]

Awesome !

[@ruqiyaarshad3745]

Is it im.shape or img shape for resizing image in your code sir?

[@gplgomes]

Does it allow to recognize my relatives names ? I tried but it only recognize famous persons.

[@yilberrojas8306]

good video, to process it into a video as input

[@다룸다림다-t8j]

Hello, thank you for making this video. Your video was so great that I subscribed to it. But I keep getting the same error. 'NoneType' object has no attribute 'shape' Even though I changed the picture, this happens, what should I do? Please give me answer🥲🥲🥲

[@임도연-b3i]

hello! we are teammate of this monkey and we had just figured it out thank u so much of your fascinating video u r sooooooooooo adorable I luv u!!!!!!!!!!!!!!!!!

[@shrikarpatil6014]

Hi , I tried this code and it is working perfectly for known faces . But I'am getting an error "'NoneType' object has no attribute 'shape' " for non matching faces , can you please help?

[@ינירעטון-ב8ג]

hi have the same problem too did you fix it?

[@muwnhius]

bill.jpg is broken

[@TechnicalGigs]

i have the same problem too did you fix it?

[@jayasmitamahapatra8140]

Very nicely explained.Can u share the link from where we can download the images

[@AdarshMenon]

colab.research.google.com/drive/15_yQKwhVsmfN2p7geco_VFGKLYqBah04?usp=sharing

[@jimmyaghera1897]

very nice work. I learned a lot from this video.

[@aqibmukhtar5217]

Bro (h, w) = img.shape[:2] pr error araha "AttributeError: 'NoneType' object has no attribute 'shape'"

[@manasisingh294]

were you able to get this fixed?

[@borisvybornykh614]

You are the best, you've helped me pass my hw

[@harshitamaurya2255]

Hey, I am getting an error 'NoneType' object has no attribute 'shape'. Please tell me how should I resolve it

[@AdarshMenon]

probably because the image was not loaded, please check if your image is in the same folder as the code / you have given the names correctly. Another reason could be there were no images detected - in this case try with a larger image or another image

[@harshitamaurya2255]

Got it...thankyou so much

[@HariHaran-qp2mw]

Know image and unknown image must be in. jpg extension

[@tommystallion7734]

@@AdarshMenon I ran your notebook without changing any code and got the same error

[@manasisingh294]

@@harshitamaurya2255 hey, how did you solve this error? Could you please share?

[@kushalappaca5324]

Can you tell what is the maximum no of classes we can consider..I'm thinking of running this for 100 masked faces????

[@haytamelbakkal825]

This is a very good video. I have a question please how can I measure the accuracy if I have a large dataset?

[@AdarshMenon]

Once you have the dataset for the correct faces to person mapping - you can count the number of correct predictions made by the face_recognition library divided by the total number of faces will be your accuracy

[@Tudismas]

why is it that the accuracy is very low for new images especially those it has no see?

[@co3mo24]

Design of Machine Learning Solution for Biometric Recognition Task with google colab do u have any idea

[@rryann088]

Thank you so much !! that Really helped understand the code ! Awesome video

[@entertainment_channel_07]

Sir can we ur own images in known and unknown folders

[@shreenejatandon32]

Code is giving error as these images are not available not and if we are giving other image's link it's not being displayed

[@gokulsai901]

Is there a way to train a haar cascade on Google colab??

[@gauravgupta4847]

Which face recognition algorithm we used in this project????

[@AdarshMenon]

based on this medium.com/@ageitgey/machine-learning-is-fun-part-4-modern-face-recognition-with-deep-learning-c3cffc121d78

[@user-rx5kq6oo9y]

@@AdarshMenon sir , in this article it is mentioned we use svm fir classifier but in implementation we dont use svm anywhere.

[@user-rx5kq6oo9y]

@@AdarshMenon does compare face uses svm at backened?

[@AdarshMenon]

@@user-rx5kq6oo9y in implementation we are using the face_recognition library which is based on the article shared

[@meghasharma154]

SIR THAT TWO KNOWN AND UNKNOWN DIRECTORY IS NOT CREATED AFTER RUN THAT DIRECTORY CAN U PLEASE TELL WHAT WE HV TO DO THIS

[@wildgoose4179]

Yea I can't get anywhere with this. I followed everything step by step and i keep getting this syntax error " File "", line 15 known_encodings = [] ^ SyntaxError: invalid syntax"

[@AdarshMenon]

Not sure what is wrong, here is the original code - colab.research.google.com/drive/15_yQKwhVsmfN2p7geco_VFGKLYqBah04?usp=sharing

[@gokulreddy2066]

Bro you are god !...You are explaining everything clearly I want to talk with you personally. Share me your linkedin in profile and let's get connected

[@AdarshMenon]

Check description

[@paramecium_sp762]

Help, i followed the video step by step but i get this error TypeError: read_img() takes 0 positional arguments but 1 was given Do you have any idea how to fix?

[@AdarshMenon]

Please check the code here - colab.research.google.com/drive/15_yQKwhVsmfN2p7geco_VFGKLYqBah04?usp=sharing read_img is a function I wrote and it takes the argument path, you probably missed it

[@paramecium_sp762]

@@AdarshMenon Thanks friend, but when i tried to run the last cell (h, w) = img.shape[:2] made an error saying AttributeError: 'NoneType' object has no attribute 'shape'

[@jacobsmith7877]

@@paramecium_sp762 Were you able to solve this?

[@ashwithathumula3283]

@@paramecium_sp762 hey did u solve this?

[@sayantandas2429]

How can I import the images in the Google collab?

[@shivendradubey4239]

Can you make a video how to do face recognition from any KZbin video

[@AdarshMenon]

I have put it on my list of videos to make. What you need to do is, download the video first (there are python libraries that take in a link and download the video), then use opencv to read the video frame by frame, then you can apply this same code on each frame, and save the output. Each frame in a video is an image

[@terrencejeffersoncimafranc100]

dude your machine is roaring. xD

[@jvsccssfdsg4500]

AttributeError Traceback (most recent call last) in () 18 19 for file in os.listdir(known_dir): ---> 20 img = read_img(known_dir + '/' + file) 21 img_enc = face_recognition.face_encodings(img)[0] 22 known_encodings.append(img_enc) in read_img(path) 6 def read_img(path): 7 img = cv2.imread(path) ----> 8 (h, w) = img.shape[:2] 9 width = 500 10 ratio = width / float(w) AttributeError: 'NoneType' object has no attribute 'shape' how to solve

[@pr0fess0rop18]

Aah yes same problem faced today morning and got a solution later now....check the file bill.jpg....its broken its infected thats why the error delete the file or get another bill photo link

[@pr0fess0rop18]

Change the Bill img address to this drive.google.com/file/d/1GuvXGjWQRe6j_Xi85kfHW67vpVk8YxQa/view?usp=sharing Also delete the 0 bytes pic from unknown

[@roshinipriyajami611]

AttributeError: 'NoneType' object has no attribute 'shape' plz help me out

[@doall321]

it is not working at import face_recognition even i install face_recognition

[@AdarshMenon]

It requires some setup to work, and so I highly recommend to use google colab

[@letsseeproductions7500]

import face_recognition line is not working in google collab

[@successlearners]

Any other libraries for face recognizing

[@kushalbhavsar3454]

You can use LBph algo from cv2

[@wallelignworkie8749]

how to change the photo, please brother this is project to present tomorrow

[@ESWARAJAYASIMHAAJAYVEMPA-yz2xo]

can we replace our images to detect

[@kushalappaca5324]

can we train multiple images using this algorithm??

[@fahrulthariq4226]

u got the answer yet?

[@puneetkaur7725]

It shows error while importing face_recognition. It says no module named face_recognition

[@pr0fess0rop18]

Pip install nahi hua hoga check it

[@mckoy1007]

why there is an error on me? AttributeError: 'NoneType' object has no attribute 'shape'

[@mckoy1007]

can someone help me?//

[@fmxy]

get a good mic, it'll dramatically improve the quality of your videos! espescially when having these screen recording type videos. blue snowball mic should be fine for the start.

[@AdarshMenon]

yeah will do

[@aietisreekar2425]

why the package is not getting installed in windows can u solve that

[@AdarshMenon]

As I have mentioned in the video, dlib and face_recognition is difficult to set up normally, and that is why I have used google colab. Check out this guide on the official repo - github.com/ageitgey/face_recognition#installing-on-windows if you want to try the setup on windows.

[@majidkhan6670]

from webcam it is possibe?

[@AdarshMenon]

technically yes, but for that you would need to set up dlib and install face_recognition library on your computer, which is not as easy as colab. See github.com/ageitgey/face_recognition

[@majidkhan6670]

@@AdarshMenon i already installed both lib...

[@AdarshMenon]

then you can try and run it

[@majidkhan6670]

@@AdarshMenon ok...but bro i need your help in my other code... i want to save a screenshot or jpg image with current time and date....iM getting error .....can u help me with tht.

[@jaisonjose8256]

macha malayali da

[@rohitkumar-nk6sd]

How to recognize images from local C drive?

[@AdarshMenon]

So since this project is on google colab, you will need to upload those images to this colab. You can find the option to do that on the top of the left side panel. Or you can also upload the images to google drive and access those files directly from colab.

[@rohitkumar-nk6sd]

@@AdarshMenon iam getting true for 2 Inputs in Boolean array what should I do please help me out

[@AdarshMenon]

@@rohitkumar-nk6sd can you share the output / more details ?

[@rohitkumar-nk6sd]

@@AdarshMenon please check your mail

[@AdarshMenon]

@@rohitkumar-nk6sd [e-mail]

[@warrior444ff7]

Sir

[@himabinduu]

Hi