Package com.banuba.sdk.recognizer

Interface Recognizer

All Known Implementing Classes:

Recognizer.CppProxy

public interface Recognizer

Nested Class Summary

Nested Classes

Modifier and Type	Interface	Description
static class	Recognizer.CppProxy

Method Summary

Modifier and Type	Method	Description
void	addFeature(Feature feature, java.util.HashSet<FeatureId> dependencies, boolean needsGpu)
void	clear()	clear input and output buffers and pipeline state, also clean cached features
static Recognizer	create(RecognizerMode mode, java.lang.Integer maxTokens)
Feature	getFeature(FeatureId feature)
static long	getFeatureId(FeatureId feature)
boolean	getNnEnable()
boolean	isDeviceNnCompatible()	Check if device compatible with NN player.
void	pause()	pause the recognizer
boolean	popFrameData(FrameData outputFrameData)
void	process(FrameData frameData)
void	pushCameraFrame(FrameData inputFrameData)
void	removeFeature(Feature feature)
void	setEyesCorrection(boolean on)	sets eyes correction
void	setFaceSearchMode(FaceSearchMode mode)	Clear pipeline state and set face search mode
void	setFeatures(java.util.HashSet<FeatureId> features)
void	setLipsCorrection(boolean on)	sets lips correction
void	setMaxFaces(int num)	sets maximum allowed face results, if face tracking feature is present
void	setNnEnable(boolean enable)	Set flag which enable NN's features.
void	setOfflineMode(boolean on)
void	start()	start the recognizer
void	stop()	pause + clear

Method Details

getFeature

@Nullable
Feature getFeature(@NonNull
FeatureId feature)

setFeatures

void setFeatures(@NonNull
java.util.HashSet<FeatureId> features)

addFeature

void addFeature(@Nullable
Feature feature,
@NonNull
java.util.HashSet<FeatureId> dependencies,
boolean needsGpu)

removeFeature

void removeFeature(@Nullable
Feature feature)

setMaxFaces

void setMaxFaces(int num)

sets maximum allowed face results, if face tracking feature is present

setOfflineMode

void setOfflineMode(boolean on)

setNnEnable

void setNnEnable(boolean enable)

Set flag which enable NN's features. True by default, if device compatible with NN player When this flag is false, recognizer will skip NN-dependent features during set_features call. In case when device have no NN player support (is_device_nn_compatible return false), setting enable to true have no any effect

getNnEnable

boolean getNnEnable()

setEyesCorrection

void setEyesCorrection(boolean on)

sets eyes correction

setLipsCorrection

void setLipsCorrection(boolean on)

sets lips correction

isDeviceNnCompatible

boolean isDeviceNnCompatible()

Check if device compatible with NN player.

setFaceSearchMode

void setFaceSearchMode(@NonNull
FaceSearchMode mode)

Clear pipeline state and set face search mode

process

void process(@Nullable
FrameData frameData)

pushCameraFrame

void pushCameraFrame(@Nullable
FrameData inputFrameData)

popFrameData

boolean popFrameData(@Nullable
FrameData outputFrameData)

start

void start()

start the recognizer

pause

void pause()

pause the recognizer

stop

void stop()

pause + clear

clear

void clear()

clear input and output buffers and pipeline state, also clean cached features

create

@Nullable
static Recognizer create(@NonNull
RecognizerMode mode,
@Nullable
java.lang.Integer maxTokens)

Parameters:

maxTokens - If greater than 0: recognizer will run processing on push-ed() frame only if it can "consume" a token - and will wait for the token to become available if all maxTokens tokens have been used up. Consumed tokens are "produced" again one-by-one when pop-ping() processed frame(s). This mode prevents processing frames more often than a consumer(e.g. renderer) can use: potentially reducing energy consumption at the cost of higher latency. Only for asynchronous mode.

getFeatureId

static long getFeatureId(@NonNull
FeatureId feature)