Sir, very good tutorial. I have a dought, if i restart the app after training the model it will not detect my face. Can you give me the steps to fix it (Edit): just added not after it will
I have a question based on the facial recognition project, the recognition for me is not done in real time, but at a certain time, for example now it recognizes, then after about 2 seconds, then after about 5, etc.. How to solve it so that it recognizes in real time and immediately detect faces
i imported the project into android studio and the app was generated successfully , I tested all features but find all features are working except face recognition, what could be the problem
How can I save permanently the captured images and use it face recognition? not in temporary even I close the apps it is no need to registered the face again?
You can save the fave vector, the numbers that we get after we run the face through the facenet model. Then we don't need to train the face again. Only scan and identify by matching against existing vectors.
Does it work offfline in android? As if employee pictures are saved once then? I want to take attendance of my 30 employee in a working room and store their today attendance to firebase realtime database.
@@themobiledev Thanks for the response but, the comparison process will happen on Cloud or Android phone? As it needed to execute ML and Vision algorithms..so will it all happen within Android Device itself?
@@themobiledevone last question, Please give me any free or paid course link that help me to make an android app via java, and make Real time attendance system with firebase database using Face Recognition. Any related course also will be helpful... Please
hello sir, great video using the code i able to do face register & recognition ,instead of the registered face here im getting the face images frm db through api which is done in webapp[face registration], while using those images the recognition is not correct, so tell me can we able to predict the face only register here, not from others, if we can, what steps i need to do after got the face image list, plz help me out sir, thank u!
am using react js in my webapp ,im capturing image like this const image = this.webcam.current.getScreenshot(); var encrypted = CryptoJS.AES.encrypt(base64, key); this.state.face1=encrypted.toString(); storing this face1 value to db- contains encrypted image, from android studio im fetching this img, decrypt & converting as bitmap this how im doing in android studio ,no errors but result is not correct -so i hope some preprocessing on image is missing ,kindly help me apart from following what is must to do in this face recognition process, plz reply sir or tell me this will work or not -! String encryptedImage= (jsonObject.get("face1")).toString(); Bitmap decryptedData=EncryptDecrypt.DecryptionAndroid(encryptedImage,AES_KEY); TensorImage tensorImage = TensorImage.fromBitmap(decryptedData); ByteBuffer faceNetByteBuffer = FaceRecognitionProcessor.faceNetImageProcessor.process(tensorImage).getBuffer(); float[][] faceOutputArray = new float[1][192]; faceNetInterpreter.run(faceNetByteBuffer, faceOutputArray);
Change this to LENS_FACING_FRONT github.com/iago001/MLSeriesDemonstrator/blob/9b70d648b32c698afd3f0a23252115b595f9c5db/app/src/main/java/com/example/mlseriesdemonstrator/helpers/MLVideoHelperActivity.java#L204
Awesome Project Sir! I am also developing a mobile application that will need this type of functionality. However, I am using Flutter. Can I replicate this functionality using flutter?
Sir, very good tutorial. I have a dought, if i restart the app after training the model it will not detect my face. Can you give me the steps to fix it
(Edit): just added not after it will
It is not designed to remember the faces. You can store the vectors in a shared preferences along with name to achieve your requirement.
@@themobiledev sir i am a beginner, i need help with your answer, and also a lot of thanks for giving us this amazing tutorial.
I have a question based on the facial recognition project, the recognition for me is not done in real time, but at a certain time, for example now it recognizes, then after about 2 seconds, then after about 5, etc.. How to solve it so that it recognizes in real time and immediately detect faces
You will have to train a better model
@@themobiledev Can you please tell me in more detail what I need to do? I am new to application programming..
Please check with the author medium.com/@estebanuri/real-time-face-recognition-with-android-tensorflow-lite-14e9c6cc53a5
i imported the project into android studio and the app was generated successfully , I tested all features but find all features are working except face recognition, what could be the problem
The models are little old, you can check if facenet has a new model. You can also try changing thr light settings while capturing pics.
How can I save permanently the captured images and use it face recognition? not in temporary even I close the apps it is no need to registered the face again?
You can save the fave vector, the numbers that we get after we run the face through the facenet model. Then we don't need to train the face again. Only scan and identify by matching against existing vectors.
@@themobiledev Sorry, but I was still confused where should I start injecting it and scan the save vector.
Does it work offfline in android? As if employee pictures are saved once then?
I want to take attendance of my 30 employee in a working room and store their today attendance to firebase realtime database.
Ya it should work offline. You can save the vectors locally and compare with them every time they are marking their attendance.
@@themobiledev Thanks for the response but, the comparison process will happen on Cloud or Android phone? As it needed to execute ML and Vision algorithms..so will it all happen within Android Device itself?
@@nisarhussainlaghari2055 on device
@@themobiledevone last question, Please give me any free or paid course link that help me to make an android app via java, and make Real time attendance system with firebase database using Face Recognition. Any related course also will be helpful... Please
@nisarhussainlaghari2055 this video has everything you nees
Thank you for this video. it works great , but in some phones, visitor analysis will not detect faces and will not show any result.
Depends on the phone, you can try reducing the video frame rate and using Accurate mode in face detection
hello sir, great video using the code i able to do face register & recognition ,instead of the registered face here im getting the face images frm db through api which is done in webapp[face registration], while using those images the recognition is not correct, so tell me can we able to predict the face only register here, not from others, if we can, what steps i need to do after got the face image list, plz help me out sir, thank u!
Just store the vector (array of numbers) in the dB. Don't store the images.
am using react js in my webapp ,im capturing image like this
const image = this.webcam.current.getScreenshot();
var encrypted = CryptoJS.AES.encrypt(base64, key);
this.state.face1=encrypted.toString();
storing this face1 value to db- contains encrypted image, from android studio im fetching this img, decrypt & converting as bitmap
this how im doing in android studio ,no errors but result is not correct -so i hope some preprocessing on image is missing ,kindly help me apart from following what is must to do in this face recognition process, plz reply sir or tell me this will work or not -!
String encryptedImage= (jsonObject.get("face1")).toString();
Bitmap decryptedData=EncryptDecrypt.DecryptionAndroid(encryptedImage,AES_KEY);
TensorImage tensorImage = TensorImage.fromBitmap(decryptedData);
ByteBuffer faceNetByteBuffer = FaceRecognitionProcessor.faceNetImageProcessor.process(tensorImage).getBuffer();
float[][] faceOutputArray = new float[1][192];
faceNetInterpreter.run(faceNetByteBuffer, faceOutputArray);
hi how can i link it to firebase and retrieve the facial values from there for recogntion
You can store the vectors in the realtime or firestore db and compare with those.
It is using back camera, what if I want to use front camera.
Change this to LENS_FACING_FRONT github.com/iago001/MLSeriesDemonstrator/blob/9b70d648b32c698afd3f0a23252115b595f9c5db/app/src/main/java/com/example/mlseriesdemonstrator/helpers/MLVideoHelperActivity.java#L204
destroy your back camera so that the system will have no choice but to use the front camera.....
Awesome Project Sir! I am also developing a mobile application that will need this type of functionality. However, I am using Flutter. Can I replicate this functionality using flutter?
The ML code will be native Android or iOS. UI and other functionality can be flutter.
Hi Sir , I also facing the same problem. did you tried this with flutter ? Can you help me please?
@@priyadharshanmanokaran4251
@@priyadharshanmanokaran4251
hi now you solve your problem please say am also face this same issue now help me
How to reuse this code with flutter?? please help me
I am not that familiar with Flutter.
Alotof thanks for your teachs
where to open this code
You will need to download Android Studio.