🧩 mlkitFaceDetection
An extension for MIT App Inventor 2.Google ML Kit Face Detection. Detects faces and returns bounding boxes, landmarks, and smile/eye details.
Specifications
Package: com.mlkit.face.detection.idriss.ma.mlkitfacedetection
Size: 2.35 MB
Version: 3.0
Minimum API Level: 21
Updated On: 2026-09-23T23:00:00Z
Built & documented using: FAST v7.1.1
Events:
mlkitFaceDetection has total 5 events.
1. FacesDetected
Fired when face detection completes. Sends a list of all detected faces as dictionaries containing bounding box, landmarks, and classifications.
| Parameter | Type |
|---|---|
| facesList | list |
2. FaceDetected
Fired for each detected face. Parameters: faceNumber (1-based index), left/top/right/bottom/width/height (bounding box), smilingProbability, leftEyeOpenProbability, rightEyeOpenProbability, and landmark coordinates.
| Parameter | Type |
|---|---|
| faceNumber | number |
| left | number |
| top | number |
| right | number |
| bottom | number |
| width | number |
| height | number |
| smilingProbability | number |
| leftEyeOpenProbability | number |
| rightEyeOpenProbability | number |
| leftEyeX | number |
| leftEyeY | number |
| rightEyeX | number |
| rightEyeY | number |
| noseX | number |
| noseY | number |
3. NoFacesDetected
Fired when the image is processed but no faces are found.
4. ErrorOccurred
Fired when an error occurs during image loading, initialization, or face detection. Check the message for details.
| Parameter | Type |
|---|---|
| message | text |
5. TestCompleted
Result of TestSetup. success=True means library, Play Services, detector, and face model are all ready. Provides status code and detailed explanation.
| Parameter | Type |
|---|---|
| success | boolean |
| status | text |
| details | text |
Methods:
mlkitFaceDetection has total 16 methods.
1. DetectFaces
Detects faces in the image at the given file path. Example: DetectFaces(/storage/emulated/0/DCIM/photo.jpg) analyzes the image and fires FacesDetected or NoFacesDetected. Supports content:// URIs, file:// paths, and Asset names. Returns immediately; results arrive via events.
| Parameter | Type |
|---|---|
| imagePath | text |
2. GetAllFaces
Returns a list of all detected faces as dictionaries. Each dictionary contains keys: faceNumber, left, top, right, bottom, width, height, smilingProbability, leftEyeOpenProbability, rightEyeOpenProbability, and landmark coordinates (leftEyeX, leftEyeY, etc.). Example: GetAllFaces() returns [{“faceNumber”: 1, “left”: 120, …}, …]
- Return type:
list
3. GetFacesCount
Returns the total number of faces detected in the last image (0 if none found).
- Return type:
number
4. HasFace
Returns True if a face exists at the given 1-based index. Example: HasFace(1) returns True if at least one face was detected.
- Return type:
boolean
| Parameter | Type |
|---|---|
| faceIndex | number |
5. GetFaceBoundingBox
Returns the bounding box as a list [left, top, right, bottom, width, height] for the face at the given 1-based index. Example: GetFaceBoundingBox(1) might return [120, 200, 400, 520, 280, 320]. Returns [-1, -1, …] if face not found.
- Return type:
list
| Parameter | Type |
|---|---|
| faceIndex | number |
6. GetFaceDetails
Returns a complete dictionary for the face at the given 1-based index. Keys include all face properties (bounding box, landmarks, classifications). Example: GetFaceDetails(1) returns {“faceNumber”: 1, “left”: 120, “smilingProbability”: 0.85, …}
- Return type:
dictionary
| Parameter | Type |
|---|---|
| faceIndex | number |
7. GetFaceLandmarks
Returns the key facial landmarks as a list [leftEyeX, leftEyeY, rightEyeX, rightEyeY, noseX, noseY] for the face at the given 1-based index. Example: GetFaceLandmarks(1) might return [150, 210, 350, 210, 250, 300].
- Return type:
list
| Parameter | Type |
|---|---|
| faceIndex | number |
8. GetBoundingBox
Returns a bounding box coordinate (left, top, right, bottom, width, height) for the face at the given 1-based index.
- Return type:
number
| Parameter | Type |
|---|---|
| faceIndex | number |
| coordinate | BoundingBoxCoordinate (helper blocks) |
- Enums for BoundingBoxCoordinate:
Left,Top,Right,Bottom,Width,Height
9. GetClassification
Returns a classification probability (smiling, left eye, or right eye) for the face at the given 1-based index.
- Return type:
number
| Parameter | Type |
|---|---|
| faceIndex | number |
| classificationType | ClassificationType (helper blocks) |
- Enums for ClassificationType:
Smiling,LeftEyeOpen,RightEyeOpen
10. GetLandmark
Returns a specific landmark coordinate (X or Y) for a facial feature for the face at the given 1-based index.
- Return type:
number
| Parameter | Type |
|---|---|
| faceIndex | number |
| landmarkType | LandmarkType (helper blocks) |
| coordinateType | CoordinateType (helper blocks) |
- Enums for LandmarkType:
LeftEye,RightEye,NoseBase,MouthLeft,MouthRight,MouthBottom - Enums for CoordinateType:
X,Y
11. IsInitialized
Returns True if the face detector is initialized and ready to process images.
- Return type:
boolean
12. IsLibraryAvailable
Returns True if the ML Kit Face Detection library classes are available on the device.
- Return type:
boolean
13. GetStatus
Returns a quick status string showing library, Google Play Services, and detector readiness. Example: “library=OK | playServices=OK | detector=READY | faces=0”. For detailed model download status, use TestSetup instead.
- Return type:
text
14. TestSetup
Performs a full diagnostic check: validates library, Google Play Services, detector initialization, and face model download status. Fires TestCompleted with results. Example: Use this button to verify everything is set up before calling DetectFaces.
15. GetVersion
Returns the version number of the ML Kit Face Detection library (e.g., “16.1.5”).
- Return type:
text
16. Close
Closes the face detector and clears all cached face data. Next DetectFaces call will reinitialize automatically.
ScreenShots
Blocks
APK
faceDetection.apk (6.4 MB)
File AIA
faceDetection.aia (2.6 MB)
AIX
com.mlkit.face.detection.idriss.ma.mlkitfacedetection.aix (2.3 MB)