tflite_flutter_helper library

TensorFlow Lite Flutter Helper Library

Makes use of TensorFlow Lite Interpreter on Flutter easier by providing simple architecture for processing and manipulating input and output of TFLite Models.

API is similar to the TensorFlow Lite Android Support Library.

Classes

BertNLClassifier
Classifier API for NLClassification tasks with Bert models, categorizes string into different classes. The API expects a Bert based TFLite model with metadata populated.
BertNLClassifierOptions
Options to configure BertNLClassifier.
BertQuestionAnswerer
Task API for BertQA models. */
BoundingBoxUtils
Helper class for converting values that represents bounding boxes into rectangles.
CastOp
Casts a TensorBuffer to a specified data type.
Category
Category is a util class, contains a label and a float value. Typically it's used as result of classification tasks.
DequantizeOp
Dequantizes a TensorBuffer with given zeroPoint and scale.
FileUtil
File I/O utilities.
FloatRingBuffer
Actual implementation of the ring buffer. */
ImageConversions
Implements some stateless image conversion methods.
ImageOperator
Operates a TensorImage object. Used in ImageProcessor.
ImageProcessor
ImageProcessor is a helper class for preprocessing and postprocessing TensorImage.
ImageProcessorBuilder
The Builder to create an ImageProcessor, which could be executed later.
LabelAxisOp
Labels TensorBuffer with axisLabels for outputs.
LabelAxisOpBuilder
Builder class to build a LabelTensor Operator.
LabelUtil
Label operation utils.
NLClassifier
Classifier API for natural language classification tasks, categorizes string into different classes.
NLClassifierOptions
Options to identify input and output tensors of the model.
NormalizeOp
Normalizes a TensorBuffer with given mean and stddev: output = (input - mean) / stddev.
Operator<T>
The common interface for classes that carries an "apply" method, which converts T to another one.
Pos
Position information of the answer relative to context. It is sortable in descending order based on logit.
Processor<T>
Processes T object with prepared Operator.
QaAnswer
Answers to QuestionAnswerer. Contains information about the answer and its relative position information to the context.
QuantizeOp
Quantizes a TensorBuffer with given zeroPoint and scale.
QuestionAnswerer
API to answer questions based on context. */
RecorderStream
ResizeOp
As a computation unit for processing images, it can resize an image to user-specified size.
ResizeWithCropOrPadOp
As a computation unit for processing images, it could resize image to predefined size.
Rot90Op
Rotates image by multiples of 90 degree.
SequentialProcessor<T>
A processor base class that chains a serial of Operator of type T and executes them.
SequentialProcessorBuilder<T>
The builder class to build a Sequential Processor.
SoundStream
SupportPreconditions
TensorAudio
TensorAudioFormat
Wraps a few constants describing the format of the incoming audio samples, namely number of channels and the sample rate. By default, channels is set to 1.
TensorBuffer
Represents the data buffer for either a model's input or its output.
TensorBufferFloat
Represents data buffer with float(double) values.
TensorBufferUint8
Represents data buffer with 8-bit unsigned integer values.
TensorImage
TensorImage is the wrapper class for Image object. When using image processing utils in Flutter Helper library, it's common to convert image objects in variant types to TensorImage at first.
TensorLabel
TensorLabel is an util wrapper for TensorBuffers with meaningful labels on an axis.
TensorOperator
Applies some operation on TensorBuffers.
TensorProcessor
TensorProcessor is a helper class for preprocessing and postprocessing tensors. It could transform a TensorBuffer to another by executing a chain of TensorOperator.
TensorProcessorBuilder
The Builder to create an TensorProcessor, which could be executed later.

Enums

BoundingBoxType
Denotes how a bounding box is represented.
CoordinateType
Denotes if the coordinates are actual pixels or relative ratios.
ResizeMethod
Algorithms for resizing.
SoundStreamStatus

Constants

methodChannel → const MethodChannel
methodChannelName → const String

Functions

enumToString(Object o) → String