App working correctly on all devices I have, but somehow on Samsung A21s, it is producing incorrect results.
It is nearly producing same output for all faces probability between .2 .3, for l2, cosine is also the same.

I thought maybe camera is the problem, so I embedded the pictures inside drawable and matched, but still it is producing nearly the same results for everything.

I noticed FloatArray returned from getCroppedFaceEmbedding producing different results for different phones.

My xiaomi mi9 works alright.

Do you have any ideas?

HI i am getting same name for every image i.e., first image name. may i know what is the problem. Please inform me is there any issue with code? If that face doesn't added in the folder also it is showing random name

I apologize for asking something so basic.

How can I save in a local text file the name of the faces that the application has detected?

Thank you very much for your help.

Hi,
Can the source of the photo not be from storage but from a photo url?
thanks

when i run the app on tablet with landscap orientation the recognation is alaways unknown

Hi, i am getting facelist.size is ''0'' . may i know why this is coming.
After that one i am getting this log I/Model: Average score for each user : {}
and then it is going into cache the error in cache was

E/Model: Exception in FrameAnalyser : null

Can you explain why this is going to cache and what i have to do to work the sample properly.

I am using using your latest code.
once the app is launched it cannot show the bounding box over the face.

Please solve my issue as possible as fast.

First of all is it supported , current version recognizes but naming is flipped for users.

Hi,

Could you please let me know how to train the facenet model ??
Any reference or links??

Thanks in advance

I really appreciated this works very well with few images. How to avoid blocking UI and also to work smoothly with many no of images ?

Hi,

I try to build the project and debug it after press USE THIS FOLDER button, suddenly the apps crash and here is the error message

` detectFacesImageByteArray.start()
Access denied finding property "ro.mediatek.platform"
detectFacesImageByteArray.end()
FATAL EXCEPTION: main
Process: com.ml.quaterion.facenetdetection, PID: 5566
java.lang.IllegalArgumentException: x must be >= 0
at android.graphics.Bitmap.checkXYSign(Bitmap.java:432)
at android.graphics.Bitmap.createBitmap(Bitmap.java:862)
at android.graphics.Bitmap.createBitmap(Bitmap.java:825)
at com.ml.quaterion.facenetdetection.BitmapUtils$Companion.cropRectFromBitmap(BitmapUtils.kt:46)
at com.ml.quaterion.facenetdetection.FileReader$getEmbedding$2.invokeSuspend(FileReader.kt:109)
at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:33)
at kotlinx.coroutines.DispatchedTask.run(DispatchedTask.kt:106)
at kotlinx.coroutines.scheduling.CoroutineScheduler.runSafely(CoroutineScheduler.kt:570)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.executeTask(CoroutineScheduler.kt:750)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.runWorker(CoroutineScheduler.kt:677)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.run(CoroutineScheduler.kt:664)
Suppressed: kotlinx.coroutines.DiagnosticCoroutineContextException: [StandaloneCoroutine{Cancelling}@4a30289, Dispatchers.Main]

Please help!

Can we adapt codes to support Landscape mode?

FATAL EXCEPTION: main
Process: com.ml.quaterion.facenetdetection, PID: 11467
java.lang.IllegalArgumentException: Internal error: Failed to run on the given Interpreter: Can not open OpenCL library on this device - dlopen failed: library "libOpenCL.so" not found
Falling back to OpenGL
TfLiteGpuDelegate Invoke: GpuDelegate must run on the same thread where it was initialized.
Node number 181 (TfLiteGpuDelegateV2) failed to invoke.
at org.tensorflow.lite.NativeInterpreterWrapper.run(Native Method)
at org.tensorflow.lite.NativeInterpreterWrapper.run(NativeInterpreterWrapper.java:249)
at org.tensorflow.lite.InterpreterImpl.runForMultipleInputsOutputs(InterpreterImpl.java:133)
at org.tensorflow.lite.Interpreter.runForMultipleInputsOutputs(Interpreter.java:80)
at org.tensorflow.lite.InterpreterImpl.run(InterpreterImpl.java:126)
at org.tensorflow.lite.Interpreter.run(Interpreter.java:80)
at com.ml.quaterion.facenetdetection.model.FaceNetModel.runFaceNet(FaceNetModel.kt:87)
at com.ml.quaterion.facenetdetection.model.FaceNetModel.getFaceEmbedding(FaceNetModel.kt:79)
at com.ml.quaterion.facenetdetection.FileReader$getEmbedding$2.invokeSuspend(FileReader.kt:108)
at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:33)
at kotlinx.coroutines.DispatchedTask.run(DispatchedTask.kt:106)
at kotlinx.coroutines.scheduling.CoroutineScheduler.runSafely(CoroutineScheduler.kt:570)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.executeTask(CoroutineScheduler.kt:750)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.runWorker(CoroutineScheduler.kt:677)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.run(CoroutineScheduler.kt:664)
Suppressed: kotlinx.coroutines.DiagnosticCoroutineContextException: [StandaloneCoroutine{Cancelling}@a0c68f4, Dispatchers.Main]

Hi, I encounter the following error when using a tflite model that has been converted and quantized by toco.

Process: com.ml.quaterion.facenetdetection, PID: 4034
java.lang.IllegalArgumentException: Cannot copy to a TensorFlowLite tensor (input) with 76800 bytes from a Java Buffer with 307200 bytes.
at org.tensorflow.lite.TensorImpl.throwIfSrcShapeIsIncompatible(TensorImpl.java:416)
at org.tensorflow.lite.TensorImpl.setTo(TensorImpl.java:140)
at org.tensorflow.lite.NativeInterpreterWrapper.run(NativeInterpreterWrapper.java:243)
at org.tensorflow.lite.InterpreterImpl.runForMultipleInputsOutputs(InterpreterImpl.java:107)
at org.tensorflow.lite.Interpreter.runForMultipleInputsOutputs(Interpreter.java:80)
at org.tensorflow.lite.InterpreterImpl.run(InterpreterImpl.java:100)
at org.tensorflow.lite.Interpreter.run(Interpreter.java:80)
at com.ml.quaterion.facenetdetection.model.FaceNetModel.runFaceNet(FaceNetModel.kt:104)
at com.ml.quaterion.facenetdetection.model.FaceNetModel.getFaceEmbedding(FaceNetModel.kt:83)
at com.ml.quaterion.facenetdetection.FileReader$getEmbedding$2.invokeSuspend(FileReader.kt:108)
at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:33)
at kotlinx.coroutines.DispatchedTask.run(DispatchedTask.kt:106)
at kotlinx.coroutines.scheduling.CoroutineScheduler.runSafely(CoroutineScheduler.kt:570)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.executeTask(CoroutineScheduler.kt:750)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.runWorker(CoroutineScheduler.kt:677)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.run(CoroutineScheduler.kt:664) 

The model is converted from facenet pretrained model to tflite by QUANTIZED_UINT8 format, the model size is reduced from 93MB to about 23MB, the problem should be caused by the mismatch between UINT8 and FLOAT, I would like to ask how to convert the tflite model in the assets directory of the project and reduce the model size while maintaining the FLOAT format.

How about implementing options for Front Camera?

Hello,

The application runs well in two phones that I tested but not in a tablet, when i try to install it show an error message "There was a problem parsing the package".

What could it be?

Hi,

Do you think doing face alignment preprocessing before send to facenet model to generate the embedding will increase the accuracy?

Model is working fine, thanks for that. But how can I switch the metric used to cosine similarity so that I can compare its performance against l2norm thanks. Btw the variable metrictobeused in frameanalyzer kt as stated in the documents does not exist.

Hi,

It was a great code thanks for it.

I am struggling from few days, but could able to achieve face recognition.

1. I have taken 2 photos of each my 7 friends. (of 720X720, I dont thing this matter much as we crop it to 160x160)
2. I launched and it processed it (Took some time to process it).
3. But the detection not working well. It is giving random names.

Is it possible to have SVM integrated with this module?

at android.os.Looper.loop(Looper.java:219)
at android.app.ActivityThread.main(ActivityThread.java:8387)
at java.lang.reflect.Method.invoke(Native Method)
at com.android.internal.os.RuntimeInit$MethodAndArgsCaller.run(RuntimeInit.java:513)
at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:1055)
Caused by: java.lang.IllegalArgumentException: Internal error: Failed to apply delegate: ModifyGraphWithDelegate is disallowed when graph is immutable.
at org.tensorflow.lite.NativeInterpreterWrapper.applyDelegate(Native Method)
at org.tensorflow.lite.NativeInterpreterWrapper.applyDelegates(NativeInterpreterWrapper.java:373)
at org.tensorflow.lite.NativeInterpreterWrapper.init(NativeInterpreterWrapper.java:85)
at org.tensorflow.lite.NativeInterpreterWrapper.(NativeInterpreterWrapper.java:63)
at org.tensorflow.lite.Interpreter.(Interpreter.java:277)
at com.ml.quaterion.facenetdetection.model.FaceNetModel.(FaceNetModel.kt:73)
at com.ml.quaterion.facenetdetection.MainActivity.onCreate(MainActivity.kt:122)
at android.app.Activity.performCreate(Activity.java:8121)
at android.app.Activity.performCreate(Activity.java:8109)

private val useGpu = true
private val useXNNPack = true
private val modelInfo = Models.FACENET_QUANTIZED

when I using rear camera for scanning face. It doesn't show bounding box overlay correctly.

App crashes on samsung mobile where as other ML KIT apps are working

I've created the images folder as you mentioned in ReadMe document but still face recognition is not happening. I'm getting following logs

2020-09-14 22:31:51.541 26060-26060/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "vendor.camera.aux.packagelist"
2020-09-14 22:31:51.550 26060-26060/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "vendor.camera.aux.packagelist"
2020-09-14 22:31:51.564 26060-26060/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "vendor.camera.aux.packagelist"
2020-09-14 22:31:51.589 26060-26093/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "persist.vendor.camera.privapp.list"
2020-09-14 22:31:51.603 26060-26078/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "vendor.camera.aux.packagelist"
2020-09-14 22:31:51.603 26060-26078/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "vendor.camera.aux.packagelist"
2020-09-14 22:31:51.722 26060-26093/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "persist.vendor.camera.privapp.list"
2020-09-14 22:31:51.733 26060-26078/com.ml.quaterion.facenetdetection E/libc: Access denied finding property "vendor.camera.aux.packagelist"
2020-09-14 22:31:52.414 26060-26212/com.ml.quaterion.facenetdetection E/Infoooooooooooooo: 90
2020-09-14 22:31:53.088 26060-26212/com.ml.quaterion.facenetdetection E/Infoooooooooooooo: 90

Hi Shubham,
I wanted to test the efficiency of facenet tf lite model so i tried running your app but it always gives the toast as " could not read images make sure you have files according to readme "

I rebuild your app and added the google services file, i added all the READ/WRITE permissions and i found that imageDir give the proper director but imageDir.listfiles return a null exception. I have directories in the images folder as specified but it is unable to list them

how to detect and recognize face when the device is using landscape orientation?

Hi everyone,

I got the error i.e.,

E/Model: Exception in FrameAnalyser ......... : Cannot copy to a TensorFlowLite tensor (input_1) with 307200 bytes from a Java Buffer with 150528 bytes.

if face is there in the frame then only i got this error.
Please help anyone to resolve this issue

I think we can migrate to Ml Kit now for this demo.
https://developers.google.com/ml-kit/migration

Hi I am trying to run it on API 21, But in FrameAnalyser.kt line no 164 boundingBoxOverlay.invalidate() will through exception.
Please help me I am a beginner Thanks in advance.

E/AndroidRuntime: FATAL EXCEPTION: Thread-3
Process: com.example.facerecognition, PID: 12921
android.view.ViewRootImpl$CalledFromWrongThreadException: Only the original thread that created a view hierarchy can touch its views.
at android.view.ViewRootImpl.checkThread(ViewRootImpl.java:7809)
at android.view.ViewRootImpl.invalidateChildInParent(ViewRootImpl.java:1338)
at android.view.ViewGroup.invalidateChild(ViewGroup.java:5446)
at android.view.View.invalidateInternal(View.java:14749)
at android.view.View.invalidate(View.java:14713)
at android.view.View.invalidate(View.java:14697)
at com.example.facerecognition.FrameAnalyser$analyze$1$1.run(FrameAnalyser.kt:150)
at java.lang.Thread.run(Thread.java:762)

I've done the following test.

I take a photograph of a person (my friend Tom) and put it in his folder. I open the app and the app reads the folder (I only have one person).

ok, the app recognizes all people like Tom, shouldn't it show a warning that his photograph is missing or directly ignore those people?

(if necessary I can record a video showing it to you)

best regards.