Package com.libremobileos.yifan.face

Class FaceRecognizer

java.lang.Object

com.libremobileos.yifan.face.FaceRecognizer

public class FaceRecognizer
extends Object

Task-specific API for detecting & recognizing faces in an image. Uses FaceFinder to detect and scan faces, FaceStorageBackend to store and retrieve the saved faces and returns the optimal result.
Refrain from using this class for registering faces into the recognition system, FaceFinder does not perform post processing and is as such better suited.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	FaceRecognizer.Face	Stores a combination of FaceScanner.Face and FaceDetector.Face, for face recognition workloads

Method Summary

Modifier and Type	Method	Description
static FaceRecognizer	create(Context ctx, FaceStorageBackend storage, float minConfidence, int inputWidth, int inputHeight, int sensorOrientation, float maxDistance, float minModelRatio)	Create FaceRecognizer instance, with matching model ratio constraint and sensible defaults regarding hardware acceleration (CPU, XNNPACK, 4 threads).
static FaceRecognizer	create(Context ctx, FaceStorageBackend storage, float minConfidence, int inputWidth, int inputHeight, int sensorOrientation, float maxDistance, float minModelRatio, boolean hwAcceleration, boolean enhancedHwAcceleration, int numThreads)	Create FaceRecognizer instance, with matching model ratio constraint.
static FaceRecognizer	create(Context ctx, FaceStorageBackend storage, float minConfidence, int inputWidth, int inputHeight, int sensorOrientation, float maxDistance, int minMatchingModels)	Create FaceRecognizer instance, with minimum matching model constraint and sensible defaults regarding hardware acceleration (CPU, XNNPACK, 4 threads).
static FaceRecognizer	create(Context ctx, FaceStorageBackend storage, float minConfidence, int inputWidth, int inputHeight, int sensorOrientation, float maxDistance, int minMatchingModels, boolean hwAcceleration, boolean enhancedHwAcceleration, int numThreads)	Create FaceRecognizer instance, with minimum matching model constraint.
List<FaceRecognizer.Face>	recognize(Bitmap input)	Detect faces and scan them

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

create

public static FaceRecognizer create(Context ctx,
                                    FaceStorageBackend storage,
                                    float minConfidence,
                                    int inputWidth,
                                    int inputHeight,
                                    int sensorOrientation,
                                    float maxDistance,
                                    int minMatchingModels,
                                    boolean hwAcceleration,
                                    boolean enhancedHwAcceleration,
                                    int numThreads)

Create FaceRecognizer instance, with minimum matching model constraint.

Parameters:

ctx - Android Context object, may be in background.

storage - The FaceStorageBackend containing faces to be recognized.

minConfidence - Minimum confidence to track a detection, must be higher than 0.0f and smaller than 1.0f

inputWidth - width of the Bitmaps that are going to be processed

inputHeight - height of the Bitmaps that are going to be processed

sensorOrientation - rotation if the image should be rotated, or 0.

maxDistance - Maximum distance (difference, not 3D distance) to a saved face to count as recognized. Must be higher than 0.0f and smaller than 1.0f

minMatchingModels - Minimum count of matching models for one face to count as recognized. If undesired, set to 1

hwAcceleration - Enable hardware acceleration (NNAPI/GPU)

enhancedHwAcceleration - if hwAcceleration is enabled, use NNAPI instead of GPU. if not, this toggles XNNPACK

numThreads - How many threads to use, if running on CPU or with XNNPACK

Returns:

FaceRecognizer instance.

See Also:

create(Context, FaceStorageBackend, float, int, int, int, float, float, boolean, boolean, int), create(Context, FaceStorageBackend, float, int, int, int, float, float), create(Context, FaceStorageBackend, float, int, int, int, float, int)

create

public static FaceRecognizer create(Context ctx,
                                    FaceStorageBackend storage,
                                    float minConfidence,
                                    int inputWidth,
                                    int inputHeight,
                                    int sensorOrientation,
                                    float maxDistance,
                                    float minModelRatio,
                                    boolean hwAcceleration,
                                    boolean enhancedHwAcceleration,
                                    int numThreads)

Create FaceRecognizer instance, with matching model ratio constraint.

Parameters:

ctx - Android Context object, may be in background.

storage - The FaceStorageBackend containing faces to be recognized.

minConfidence - Minimum confidence to track a detection, must be higher than 0.0f and smaller than 1.0f

inputWidth - width of the Bitmaps that are going to be processed

inputHeight - height of the Bitmaps that are going to be processed

sensorOrientation - rotation if the image should be rotated, or 0.

maxDistance - Maximum distance (difference, not 3D distance) to a saved face to count as recognized. Must be higher than 0.0f and smaller than 1.0f

minModelRatio - Minimum count of matching models for one face to count as recognized. Must be higher or equal to 0.0f and smaller or equal to 1.0f. If undesired, set to 0f

hwAcceleration - Enable hardware acceleration (NNAPI/GPU)

enhancedHwAcceleration - if hwAcceleration is enabled, use NNAPI instead of GPU. if not, this toggles XNNPACK

numThreads - How many threads to use, if running on CPU or with XNNPACK

Returns:

FaceRecognizer instance.

See Also:

create(Context, FaceStorageBackend, float, int, int, int, float, int, boolean, boolean, int), create(Context, FaceStorageBackend, float, int, int, int, float, float), create(Context, FaceStorageBackend, float, int, int, int, float, int)

create

public static FaceRecognizer create(Context ctx,
                                    FaceStorageBackend storage,
                                    float minConfidence,
                                    int inputWidth,
                                    int inputHeight,
                                    int sensorOrientation,
                                    float maxDistance,
                                    int minMatchingModels)

Create FaceRecognizer instance, with minimum matching model constraint and sensible defaults regarding hardware acceleration (CPU, XNNPACK, 4 threads).

Parameters:

ctx - Android Context object, may be in background.

storage - The FaceStorageBackend containing faces to be recognized.

minConfidence - Minimum confidence to track a detection, must be higher than 0.0f and smaller than 1.0f

inputWidth - width of the Bitmaps that are going to be processed

inputHeight - height of the Bitmaps that are going to be processed

sensorOrientation - rotation if the image should be rotated, or 0.

maxDistance - Maximum distance (difference, not 3D distance) to a saved face to count as recognized. Must be higher than 0.0f and smaller than 1.0f

minMatchingModels - Minimum count of matching models for one face to count as recognized. If undesired, set to 1

Returns:

FaceRecognizer instance.

See Also:

create(Context, FaceStorageBackend, float, int, int, int, float, float, boolean, boolean, int), create(Context, FaceStorageBackend, float, int, int, int, float, int, boolean, boolean, int), create(Context, FaceStorageBackend, float, int, int, int, float, float)

create

public static FaceRecognizer create(Context ctx,
                                    FaceStorageBackend storage,
                                    float minConfidence,
                                    int inputWidth,
                                    int inputHeight,
                                    int sensorOrientation,
                                    float maxDistance,
                                    float minModelRatio)

Create FaceRecognizer instance, with matching model ratio constraint and sensible defaults regarding hardware acceleration (CPU, XNNPACK, 4 threads).

Parameters:

ctx - Android Context object, may be in background.

storage - The FaceStorageBackend containing faces to be recognized.

minConfidence - Minimum confidence to track a detection, must be higher than 0.0f and smaller than 1.0f

inputWidth - width of the Bitmaps that are going to be processed

inputHeight - height of the Bitmaps that are going to be processed

sensorOrientation - rotation if the image should be rotated, or 0.

maxDistance - Maximum distance (difference, not 3D distance) to a saved face to count as recognized. Must be higher than 0.0f and smaller than 1.0f

minModelRatio - Minimum count of matching models for one face to count as recognized. Must be higher or equal to 0.0f and smaller or equal to 1.0f. If undesired, set to 0f

Returns:

FaceRecognizer instance.

See Also:

create(Context, FaceStorageBackend, float, int, int, int, float, int, boolean, boolean, int), create(Context, FaceStorageBackend, float, int, int, int, float, float, boolean, boolean, int), create(Context, FaceStorageBackend, float, int, int, int, float, int)

recognize

public List<FaceRecognizer.Face> recognize(Bitmap input)

Detect faces and scan them

Parameters:

input - Bitmap to process

Returns:

List of FaceRecognizer.Faces