Explore Firebase MLKit for Android

My tech talk on Explore Firebase on ML Kit for Android and GDG DevFest At Ahmedabad.

Transcript

1. Hiren Dave @hjdave Exploring Firebase ML Kit on Android
2. What are the skills required for Machine Learning? Ahmedabad
3. End Skills Required for ML Develop a passion for learning Supervised Learning Unsupervised Learning Linear Regression Neural Networks Naïve Byes Logistic Regression TensorFlow Numpy / Pandas Python
4. “To think creatively, we must be able to look afresh at what we normally take for granted” - George Keller Hey, I am a mobile app developer. Don’t you think it’s too much learning for me?
5. No Problem. We have a solution. Introducing ML Kit Ahmedabad
6. Introduction to ML Kit for Firebase At Google I/O 2018, ML Kit was introduced as part of Firebase suite. The SDK brings power of Google Vision APIs, TensorFlow Lite and Neural Network APIs together in single SDK. To use the ML Kit, you don’t have to be skilled ML developer
7. What is ML Kit for Firebase? Ahmedabad
8. Ready To Use APIS It comes with set of prebuilt ready to use APIs for ML use cases like Text recognition Face detection Barcode scanning Image labeling Shape detection Pre built APIs for mobile apps
9. How to use pre built APIs of ML Kit in mobile app. Ahmedabad
10. Integrate SDK Prepare Input Apply ML Model Deploy Chart Data Source Info Use APIs in Android Mobile App
11. Getting Started With ML Kit in Android App Ahmedabad
12. Connect to Firebase • Click Tools > Firebase • Select Connect Your App To Firebase • Click on Connect to Firebase • Add displayed code to your app. That’s it. Your app is now connected to Firebase
13. Add ML Kit Dependencies Add following dependencies in your gradle implementation 'com.google.firebase:firebase-ml-vision:16.0.0' implementation 'com.google.firebase:firebase-ml-vision-image-label-model:15.0.0'
14. Add Extra Dependencies Add following dependencies in your manifest file. <meta-data android:name="com.google.firebase.ml.vision.DEPENDENCIES" android:value="barcode, face,ocr" />
15. What all you can do with ML Kit Ahmedabad
16. Barcode Scanning Add barcode scanning to your app using ML Kit
17. Face Detection Add face detection to your app using ML Kit
18. Text Recognition Add text recognition to your app using ML Kit
19. Import Packages import com.google.firebase.ml.vision.FirebaseVision import com.google.firebase.ml.vision.common.FirebaseVisionImage import com.google.firebase.ml.vision.face.FirebaseVisionFace import com.google.firebase.ml.vision.face.FirebaseVisionFaceDetectorOptions import com.google.firebase.ml.vision.face.FirebaseVisionFaceLandmark
20. Setup Image Detector Options val firebaseVisionImage = FirebaseVisionImage.fromBitmap(image) val options = FirebaseVisionFaceDetectorOptions.Builder() .setModeType(FirebaseVisionFaceDetectorOptions.ACCURATE_MODE .setLandmarkType(FirebaseVisionFaceDetectorOptions.ALL_LANDMARKS) .setClassificationType(FirebaseVisionFaceDetectorOptions.ALL_CLASSIFIC ATIONS) .setMinFaceSize(0.15f) .setTrackingEnabled(true) .build()
21. Create detector detector = FirebaseVision.getInstance().getVisionFaceDetector(options); detector.detectInImage(firebaseVisionImage) .addOnSuccessListener { processFace(it) } .addOnFailureListener { Log.e("error", it.message) }
22. Create Text Detector val detector = FirebaseVision.getInstance().visionTextDetector detector.detectInImage(textImage) .addOnCompleteListener { // More code here }
23. On Cloud APIs Cloud APIs leverage the power of Google Cloud Platform’s ML Technology ML Kit also gives you cloud APIs.
24. Setup Firebase Project • Login to Firebase console • Create project • Upgrade to Blaze Plan • Enable Cloud Based APIs • Copy API Key in your code That’s it and now you can work with cloud base API without adding dependencies in your app.
26. How to use Cloud APIs Ahmedabad
27. Add Dependencies compile 'com.google.api-client:google-api-client-android:1.22.0' compile 'com.google.apis:google-api-services-vision:v1-rev357-1.22.0' compile 'com.google.code.findbugs:jsr305:2.0.1'
28. Configure API Client and Feature Vision.Builder visionBuilder = new Vision.Builder( new NetHttpTransport(), new AndroidJsonFactory(), null); visionBuilder.setVisionRequestInitializer( new VisionRequestInitializer("YOUR_API_KEY")); Feature desiredFeature = new Feature(); desiredFeature.setType("FACE_DETECTION");
29. Use Custom Model If you want to use existing ML models in your mobile app, you can use it along with ML Kit Use custom TensorFlow Lite model
30. Use Custom TensorFlow Model • Generate quality dataset. • Train the model • Upload Model to Firebase • Navigate to ML Kit in Firebase • Select custom model and upload model • Load the Model • Run the model over data Now you can use custom model in your mobile app.
32. Demo – Use TensorFlow Model Ahmedabad
33. Hiren Dave @hjdave Thank you!
34. Q&A