search

shubham0204 / FaceRecognition_With_FaceNet_Android

Face Recognition using the FaceNet model and MLKit on Android.
https://towardsdatascience.com/using-facenet-for-on-device-face-recognition-with-android-f84e36e19761
Apache License 2.0
252 stars 88 forks source link

App crash on samsung #1

Closed raja259 closed 3 years ago

raja259 commented 4 years ago

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

shubham0204 commented 4 years ago

Can you please include your Android version? If you have built the Android project and then you are running the app on your device, make sure you include the crash logs.

handhikadj commented 4 years ago

it occurs on mine also. it crashes on my A20 on Android 10

shubham0204 commented 4 years ago

@handhikadj Can you provide crash logs for the same? Also, let me know if you have placed images ( of people you want to recognize ) in an images folder in your device's internal storage.

handhikadj commented 4 years ago

I haven't. but honestly it's annoying that we should use firebase to make it working. can you make it 100% offline?

shubham0204 commented 4 years ago

Firebase MLKit gives us the bounding boxes for all the faces present in the image. We crop these faces and feed them to our FaceNet model. The FaceNet model works best on cropped images ( as unwanted information is discarded by cropping the image ).

handhikadj commented 4 years ago

i followed this (official) sample: https://github.com/googlesamples/mlkit/tree/master/android/vision-quickstart/app/src/main/java/com/google/mlkit/vision/demo the face detection works flawlessly. it provides bounding box, 100% offline and follows the latest technology and API Android has to offer but the problem is I still don't know how to integrate with tensorflow lite. would you check it out?

shubham0204 commented 4 years ago

@handhikadj , the Android project included in this repo uses the same Firebase MLKit API as the official samples do. The project works offline just as the official sample does.

shubham0204 commented 4 years ago

Do you have any other details regarding this issue? For proposing new features, consider adding a pull request. If you don't have any further details regarding the issue, it would be worth to close this issue.

shubham0204 commented 4 years ago

I have fixed the bug in the latest commit. Please see if the problem is solved.

shubham0204 commented 4 years ago

As no further details are provided by the participants regarding the latest commit, I am closing this issue for time being.