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

Types

EfficientDetObjectDetectionModel

jvm

class EfficientDetObjectDetectionModel(internalModel: OnnxInferenceModel, modelKindDescription: String?) : EfficientDetObjectDetectionModelBase<BufferedImage> , InferenceModel

Special model class for detection objects on images with built-in preprocessing and post-processing.

EfficientDetObjectDetectionModelBase

common

abstract class EfficientDetObjectDetectionModelBase(modelType: String?) : ObjectDetectionModelBase

Base class for object detection models based on EfficientDet architecture.

ObjectDetectionModelBase

common

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

Base class for object detection models.

SSDLikeModel

android

class SSDLikeModel(internalModel: OnnxInferenceModel, metadata: SSDLikeModelMetadata, modelKindDescription: String?) : SSDLikeModelBase<Bitmap> , CameraXCompatibleModel, InferenceModel

Special model class for detection objects on images with built-in preprocessing and post-processing. Suitable for models with SSD like output decoding.

SSDLikeModelBase

common

abstract class SSDLikeModelBase(metadata: SSDLikeModelMetadata, modelType: String?) : ObjectDetectionModelBase

Base class for object detection model based on SSD architecture.

SSDLikeModelMetadata

common

data class SSDLikeModelMetadata(outputBoxesName: String, outputClassesName: String, outputScoresName: String, yMinIdx: Int, xMinIdx: Int)

This class aggregates the metadata of the SSD-like model used for decoding the output. The class is exists mostly for reducing code duplication.

SSDMobileNetV1ObjectDetectionModel

jvm

class SSDMobileNetV1ObjectDetectionModel(internalModel: OnnxInferenceModel, modelKindDescription: String?) : SSDLikeModelBase<BufferedImage> , InferenceModel

Special model class for detection objects on images with built-in preprocessing and post-processing.

SSDObjectDetectionModel

jvm

class SSDObjectDetectionModel(internalModel: OnnxInferenceModel, modelKindDescription: String?) : SSDLikeModelBase<BufferedImage> , InferenceModel

Special model class for detection objects on images with built-in preprocessing and post-processing.

Functions

detectObjects

android

fun ObjectDetectionModelBase<Bitmap>.detectObjects(imageProxy: ImageProxy, topK: Int = 3): List<DetectedObject>

Returns the detected object for the given image sorted by the score. Internal preprocessing is updated to rotate image to match target orientation. After prediction, internal preprocessing is restored to the original state.