amazon-chime-sdk / com.amazonaws.services.chime.sdk.meetings.audiovideo.video.backgroundfilter / SegmentationProcessor

SegmentationProcessor

class SegmentationProcessor

SegmentationProcessor predicts foreground mask for an image.

Constructors

<init>

SegmentationProcessor predicts foreground mask for an image.

SegmentationProcessor(context: Context)

Properties

context

val context: Context

modelState

lateinit var modelState: ModelState

Functions

getInputBuffer

Retrieve input image on which segmentation can be applied.

fun getInputBuffer(): ByteBuffer

getOutputBuffer

Retrieve segmented image mask.

fun getOutputBuffer(): ByteBuffer

initialize

Initialize and load the tensorflow model.

fun initialize(width: Int, height: Int, modelShape: ModelShape): Unit

predict

Predicts the foreground on the input image.

fun predict(): PredictResult

Companion Object Properties

isMachineLearningLibraryLoaded

var isMachineLearningLibraryLoaded: Boolean