org.jetbrains.kotlinx.dl.onnx.inference.posedetection

MultiPoseDetectionModel

jvm

class MultiPoseDetectionModel(internalModel: OnnxInferenceModel, modelKindDescription: String?) : MultiPoseDetectionModelBase<BufferedImage> , InferenceModel

MultiPoseDetectionModel is an ultra-fast and accurate model that detects 6 persons with 17 keypoints and 18 basic edges of a body for each of them.

MultiPoseDetectionModelBase

common

abstract class MultiPoseDetectionModelBase<I>(modelKindDescription: String?) : OnnxHighLevelModel<I, <ERROR CLASS>>

Base class for pose detection models for detecting multiple poses per image.

SinglePoseDetectionModel

class SinglePoseDetectionModel(internalModel: OnnxInferenceModel, modelKindDescription: String?) : SinglePoseDetectionModelBase<Bitmap> , InferenceModel, CameraXCompatibleModel

class SinglePoseDetectionModel(internalModel: OnnxInferenceModel, modelKindDescription: String?) : SinglePoseDetectionModelBase<BufferedImage> , InferenceModel

SinglePoseDetectionModel is an ultra-fast and accurate model that detects 17 keypoints and 18 basic edges of a body.

SinglePoseDetectionModelBase

common

abstract class SinglePoseDetectionModelBase<I>(modelKindDescription: String?) : OnnxHighLevelModel<I, <ERROR CLASS>>

Base class for pose detection models for detecting a single pose per image.

detectPose

android

fun SinglePoseDetectionModelBase<Bitmap>.detectPose(imageProxy: ImageProxy): DetectedPose

Detects a pose for the given imageProxy. Internal preprocessing is updated to rotate image to match target orientation. After prediction, internal preprocessing is restored to the original state.

edgeKeyPointsPairs

common

val edgeKeyPointsPairs: List<<ERROR CLASS><Int, Int>>

Pair of points which define body edges.

keyPoints

common

val keyPoints: Map<Int, String>

Dictionary that maps from joint names to keypoint indices.

Package org.jetbrains.kotlinx.dl.onnx.inference.posedetection

Types

Functions

Properties