Face Recognition Using Keras And OpenCV- Part 1- Data Collection

Рет қаралды 70,235

4 жыл бұрын

Please join as a member in my channel to get additional benefits like materials in Data Science, live streaming for Members and many more
/ @krishnaik06
Code-Github
github.com/krishnaik06/Deep-L...
Please do subscribe my other channel too
/ @krishnaikhindi
Connect with me here:
Twitter: / krishnaik06
Facebook: / krishnaik06
instagram: / krishnaik06

Пікірлер: 95

@gaurishankaracharya7652 3 жыл бұрын

Greate channel everything you can find about related to Datascience. definitely I will recommend this channel to my friends who want to learn Data Science. Great job sir🙏

@shirjeelkhan 4 жыл бұрын

Thank you sir, please continue this series.

@mayurikadtan1076 4 жыл бұрын

Hello krish.. Really very nice video.. In addition to this Can you make one video on handwritten signatures recognition using siamese network from scratch .. Thanking you in anticipation

@SivaKumar-rg1kx 4 жыл бұрын

I joined in ineuron for machine learning training. Is these project videos/concepts cover in that training?

@hamzataoujouti3562 3 жыл бұрын

the best explanation ever gj

@sibamarcel9428 4 жыл бұрын

You are doing amazin Job...

@nidhirbhavsar3917 3 жыл бұрын

Thanks a lot man this was really helpful

@wimachannel4702 3 жыл бұрын

Hello, excellent explanation. But I have a question it is possible to do facial recognition with masked faces? We can use eyes detection or another method?

@anusmirza6324 4 жыл бұрын

Thank you very much sir.please upload next part.waiting.

@sivaramarajuk9993 4 жыл бұрын

Good initiative sir .....👍👍👍

@mandarkulkarni823 4 жыл бұрын

Hi,@krish naik,is there any other mode of payment...bcoz i am not able to make payment may be bank servers are down due to lockdown..plz reply aeap...

@shripradgaydole6159 3 жыл бұрын

Hello Sir can you create some tutorials on one-shot learning using siamese neural network or any other which is best suitable for face recognition.

@ojasviverma7055 2 жыл бұрын

Sir I have images with different background objects so I am getting errors in prediction, so is it necessary to have plain background?

@josephadamu2723 3 жыл бұрын

Please can you suggest some articles relating to this project

@subhajitchoudhury4054 3 жыл бұрын

Sir, What's the system configuration for running such codes??

@shivant165 4 жыл бұрын

Sir please make vedio on data mining algorithms like aprior ect......

@ashwinraj8918 4 жыл бұрын

Sir, please upload the next part of the face recognition.

@shripradgaydole6159 4 жыл бұрын

Sir can i just change the model to SSD mobilnet here will that work better ?

@radhikapatil8003 4 жыл бұрын

Hi sir can it be implimented using pytorch??? please explain

@AbhishekKumar-zl7rq 4 жыл бұрын

if you get error like this TypeError: int() argument must be a string, a bytes-like object or a number, not 'NoneType' this is due to full path is properly added eg: add from 'C:/Users/Madhu Devara/Downloads/opencv/faces' to 'C:/Users/Madhu Devara/Downloads/opencv/faces/' see i added "/" at last

@shareknowledge7125 3 жыл бұрын

Sir please tell which play list , This vedio is there. I am search deep learning play list and not able to find. Can you help me out

@gauravbhoi6627 3 жыл бұрын

If I add more and more people over time do i have to train model again and again

@mihretdesta9153 Жыл бұрын

hey sir what is the reason for increase the training loss and validation loss

@princevegeta3707 3 жыл бұрын

How many images should we need to train

@vinayakparab6260 3 жыл бұрын

Hello Sir. Thanks for this amazing tutorial. I ran the training phase of the model on Google colab and I am getting accuracy of 100 % on the training data during the 4th epoch. How do I tune the parameters so that the data does not overfit to the model ? I am running the model for 5 epochs

@teume9788 2 жыл бұрын

how did you run on google colab can you please help?

@user-wt9ug3mb8g 3 жыл бұрын

Can we use SQL instead of folders?

@AKASHDEEP-xs1bo 4 жыл бұрын

Can I implement using pi 4b

@ruthwiksaivenigandla1280 3 жыл бұрын

sir..can i use this code and setup for my mini project?

@radhikapatil8003 4 жыл бұрын

sir please make a video on YOLO FACE DETECTION ,THANK YOU

@aditithakur9281 Жыл бұрын

i need help regarding this I have made face recognition attendance management...but it is not 💯 present accurate.....it is also mixing data ...help me please 😭 working correctly butt after using again and again it keeps mixing data help me please

@adityataksande5024 4 жыл бұрын

Thank you Sir , i have a question for you how to make machine learning model for deployment ?? please answer it soon.

@adityataksande5024 4 жыл бұрын

@Viral 10 no please tell me how to convert ml model using pickle?

@arjyabasu1311 4 жыл бұрын

After i run the coding , i get error like "Unable to open file (unable to open file: name = 'facefeatures_new_model_final.h5". how to resolve this ?

@mychanalmyproperty1810 4 жыл бұрын

You have downloaded it and places it it particular folder ...as path you trying open that file

@sunidhigupta5079 4 жыл бұрын

detection of any sort of thing from neural network in live video feed #resnet50 #face mask detection #counting no of people #distance between persons in a room Plzz make a video on this project

@rohithreddypeddi4148 3 жыл бұрын

where the photos from code are saved?

@theseriousguy2136 3 жыл бұрын

I have one doubt that....face recognition is considered to be one shot learning problem.......so shouldn't have you collected at-max 2-3 images of a person....rather than 200-300 images of a person..

@rohitgupta-ts2jf 3 жыл бұрын

R u making use of CPU or GPU ? N if u r using GPU then please help me with ... cause m not able to use my gpu .!!

@shareknowledge7125 3 жыл бұрын

Thank you sir

@karnavchinchavde4851 Жыл бұрын

Can we use the same code for mobilenets v2?

@alphonseinbaraj7602 4 жыл бұрын

Sir , After i run the coding , i get error like "Unable to open file (unable to open file: name = 'facefeatures_new_model_final.h5". how to resolve this ? please

@arjyabasu1311 4 жыл бұрын

Same problem bro!!

@bhaveshsalvi4437 4 жыл бұрын

@@arjyabasu1311 cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml') try this

@abdulqayoom-abro 3 жыл бұрын

boss need your help to build real time face tracking system for institute using python

@fahadpatel5731 4 жыл бұрын

Sir plzzz upload the next part fast Plzzz🙏🙏, it’s a humble request

@darpan810 4 жыл бұрын

I don't have high configured laptop Can I use google collaboratory for this project

@alphonseinbaraj7602 4 жыл бұрын

hi .you can do this using colab .

@ram54321 3 жыл бұрын

why are u using the word scratch at the beginning of video?@Krish Naik

@EfeUNLU98 3 жыл бұрын

i did it, however, the camera is opening and shut down immediately, then I see the "Face not found" script. why?

@Vinay-jg7yb 3 жыл бұрын

because this code is bullshit

@vincentedeh3926 Жыл бұрын

Please can someone send me the full code. Its no longer on github

@devikams6479 4 жыл бұрын

when I run it in jupiter notebook ,no window is popping up to capture my images.Can someone help?

@anilkumar-cu6we 3 жыл бұрын

yes it will pop up just check it one more time!

@EfeUNLU98 3 жыл бұрын

@@anilkumar-cu6we bana ulasabilir misin

@EfeUNLU98 3 жыл бұрын

@@anilkumar-cu6we benim hala 1 kere acılıyo sonra face not found

@aryashah9073 4 жыл бұрын

Hello brother. How do you downloaded the photos from google images. I couldn’t download more than 100. Can you help me ??

@nithinreddy2299 3 жыл бұрын

He didn't download those images. He captured those images by running the code and camera will open and capture the screen.

@aryashah9073 3 жыл бұрын

Nithin Reddy Can we download random images from google images??

@nithinreddy2299 3 жыл бұрын

@@aryashah9073 you can try with your images and your friends images or else you can download any famous celebrity images

@rajasekarsuriya6269 3 жыл бұрын

@@nithinreddy2299 He have saved his friend's images in training data.What images he has saved in test data??

@saikatsom8092 2 жыл бұрын

can it handle spoofing ?

@rajeshchowdhury2505 4 жыл бұрын

Keras Resnet instead of MobileNet upload a video

@AyazurRahman 4 жыл бұрын

I tried this but it saves only one picture in Images folder and It gives message "Face not found" in Jupyter as output. I have followed your instruction step to step, I don't understand what am i doing wrong?

@bhaveshsalvi4437 4 жыл бұрын

cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml') Try this

@salmashaik03 4 жыл бұрын

Did u get it after trying that code??

@salmashaik03 4 жыл бұрын

@@bhaveshsalvi4437 I tried this but still it gives a message face not found in jupyter

@AyazurRahman 4 жыл бұрын

salma shaik Sorry I missed that notification. Didn't try it ☹️

@prithikasoundararajan148 3 жыл бұрын

@@AyazurRahman I am getting 'Face not found' message continuously, please help resolve

@randomposts8837 3 жыл бұрын

code is running but i unable to find the images folder can someone tell me what will be the issue?

@adityan8536 3 жыл бұрын

you have to create a images folder first and then run the code

@UsamaChishti 4 жыл бұрын

please tell me that how to insert data in test folder?

@Vinay-jg7yb 3 жыл бұрын

did you find out how it can be done?

@UsamaChishti 3 жыл бұрын

@@Vinay-jg7yb I have used the python OS library for creating directories. Get a name from the user and create a directory with its name and then give the path where you want to store it.

@Vinay-jg7yb 3 жыл бұрын

@@UsamaChishti thanks for replying What will be there in test folder?

@devikams6479 4 жыл бұрын

face cropper is not coming

@saptarshipaul6636 2 жыл бұрын

This code is not working

@zlq8807 3 жыл бұрын

good

@kundansinha5603 3 жыл бұрын

I'm using the same code, however, it's taking ages to capture images. I am getting a lot of "Face not found". Is there a specific reason (using laptop cam)? Sometimes it takes a minute to capture 100 pics as face_extractor return value is blank/none. Thanks

@EfeUNLU98 3 жыл бұрын

did you solve the problem

@kundansinha5603 3 жыл бұрын

@@EfeUNLU98 yes, all you need to do is to tweak the face_classifier parameters. Try faces = face_classifier.detectMultiScale(frame, 1.1, 4) should work cheers