@awarns/ml-kit
v1.0.1
Published
Run CNN and MLP machine learning models for regression and classification tasks in your smartphone
Downloads
3
Readme
@awarns/ml-kit
This module allows to execute TensorFlow Lite machine learning models embedded in the mobile device, which is useful to make predictions based on some input data.
Supported models architectures are Convolutional Neural Networks (CNN) and Multilayer Perceptrons (MLP), both for classification and regression problems.
Install the plugin using the following command line instruction:
ns plugin add @awarns/ml-kit
Usage
After installing this plugin, you will have access to an API and two tasks to perform classification or regression on the provided input data. But before using the plugin, you must meet the following requirements regarding the machine learning model:
- Have a TensorFlow Lite machine learning model (*.tflite) of the supported architectures (CNN or MLP). Classification models must have metadata (i.e., name, version, author, etc...) and an associated labels file with each label in a row of the file. While regression models don't have to include an associated labels file, it's recommended to add the metadata with the information of the model.
- Place your TensorFlow Lite models in a folder named ml-models inside your app's src folder (i.e., same level as {app|main}.ts).
- The model file name must follow the next format: {model_name}-{cnn|mlp}-[version].tflite. The file name must contain a name (model_name), the model's architecture (cnn or mlp) and, optionally, the model's version (version). The file name elements must be splitted by a dash (-).
API
In order to do a regression or a classification, first you have to load a machine learning model. You can do that using the getModel(...)
method provided
by the ModelManager
. It also provides the listModels
method, useful known which models are available in the device.
When you load a model using the getModel(...)
method, you obtain a BaseModel
or a ClassificationModel
to perform a regression or a
classification, respectively. To do so, you can create a Regressor
using a BaseModel
, and a Classifier
using
a ClassificationModel
. Finally, to perform the prediction, you just have to call to the predict
method, which will return a RegressionResult
or a ClassificationResult
, depending on which predictor has been used.
Here's a complete example:
import {
BaseModel,
ClassificationModel,
ClassificationResult,
Classifier,
DelegateType,
getModelManager,
InputData,
ModelType, RegressionResult, Regressor
} from '@awarns/ml-kit';
async function doClassification(inputData: InputData /* number[] */) {
const model: ClassificationModel = await getModelManager().getModel(
'activity_classifier-cnn',
ModelType.CLASSIFICATION,
{ acceleration: DelegateType.GPU } // Use GPU, if available.
);
const classifier = new Classifier(model);
const result: ClassificationResult = classifier.predict(inputData);
}
async function doRegression(inputData: InputData /* number[] */) {
const model: BaseModel = await getModelManager().getModel(
'stress_regressor-cnn',
ModelType.REGRESSION,
{ acceleration: 4 } // Use 4 threads.
);
const regressor = new Regressor(model);
const result: RegressionResult = regressor.predict(inputData);
}
Note: the
RegressionResult
andClassificationResult
are not framework records. If you want to introduce these results into the framework for example, to persist them using the persistence package, you have to manually create aRegression
andClassification
records.
ModelManager
You can obtain the singleton instance of the ModelManager
calling the getModelManager()
function.
| Method | Return type | Description |
|----------------------------------------------------------------------------------|----------------------------------------------------------|----------------------------------------------------------------|
| listModels()
| Promise<Model[]>
| Returns a list of the models that are available for their use. |
| getModel(modelName: string, modelType: ModelType, modelOptions?: ModelOptions)
| Promise<BaseModel|ClassificationModel> | Retrieves and loads the specified model, ready to be used. |
Model
| Property | Type | Description |
|-------------|-------------|----------------------------|
| modelInfo
| ModelInfo
| Contains model's metadata. |
ModelInfo
| Property | Type | Description |
|----------------|---------------------|----------------------------------------------------------|
| id
| string
| Identifier of the model, generally the name of its file. |
| name
| string
| Name info included in the model's metadata. |
| architecture
| ModelArchitecture
| Architecture of the model, i.e., CNN
or MLP
. |
| version
| string
| Version info included in the model's metadata. |
| author
| string
| Author info included in the model's metadata. |
ModelType
| Value | Description |
|------------------|-----------------------------------------|
| REGRESSION
| A model that performs a regression. |
| CLASSIFICATION
| A model that performs a classification. |
ModelOptions
| Property | Type | Description |
|----------------|-----------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| acceleration
| DelegateType | number | Which type of acceleration to use when running the model. It can take the values DelegateType.GPU
(GPU acceleration), DelegateType.NNAPI
(Android Neural Networks API acceleration) or a number indicating the quantity of threads to use. |
Regressor
| Method | Return type | Description |
|---------------------------------|-----------------------------------------|------------------------------------------------|
| predict(inputData: InputData)
| RegressionResult
| Preforms a regression using the provided data. |
Classifier
| Method | Return type | Description |
|---------------------------------|-------------------------------------------------|----------------------------------------------------|
| predict(inputData: InputData)
| ClassificationResult
| Preforms a classification using the provided data. |
Tasks
| Task name | Description |
|-------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| {classificationAim}Classification{tag?}
| It performs a classification using the input data contained on the invocation event's payload. classificationAim
is used to differentiate among classification tasks. An optional tag
can be added to the task name. |
| {regressionAim}Regression{tag?}
| It performs a regression using the input data contained on the invocation event's payload. regressionAim
is used to differentiate among regression tasks. An optional tag
can be added to the task name. |
Note: the input data provided through the invocation event's payload must be an array of numbers ready to be feed into the model. In other words, the main application is the one in charge of executing the required data preprocessing techniques (e.g., normalization, feature extraction, etc...) to prepare the data for the model, not this module.
To register these tasks for their use, you just need to import them and call their generator functions inside your application's task list:
import { Task } from '@awarns/core/tasks';
import {
classificationTask,
regressionTask,
DelegateType,
} from '@awarns/ml-kit';
import { DelegateType } from './index';
export const demoTasks: Array<Task> = [
classificationTask('human-activity', 'activity_classifier-mlp', '', { acceleration: 4 }),
// humanActivityClassification
classificationTask('human-activity', 'activity_classifier-cnn', 'UsingCNN', { acceleration: DelegateType.GPU }),
// humanActivityClassificationUsingCNN
regressionTask('stress-level', 'stress_regressor-cnn'),
// stressLevelRegression
]
Task generator parameters:
| Parameter name | Type | Description |
|--------------------------------------------------|-------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| {classification|regression}Aim | string
| Objective of the classification/regression. Used to name the task. |
| modelName
| string | ModelNameResolver | Name of the model (without tflite extension) stored in the ml-models folder to use for this task, or a function that returns the name of the model when called. |
| tag
(Optional) | string
| Adds a tag to the name of the task to differentiate it from other tasks with other configurations. |
| modelOptions
(Optional) | ModelOptions | ModelOptionsResolver | Configuration to use with the model or a function that returns the configuration when called. |
Note: It's highly recommended to provide the
{classification|regression}Aim
in snake-case format. This string will be used as the type of the output record of the task. All the records have their type in snake-case, so providing the{classification|regression}Aim
in snake-case will keep the consistency of the framework.
ModelNameResolver: () => string
Useful to change the model used by a task at runtime. You can use the ModelManager
to obtain a list with the models that are available in the device.
ModelOptionsResolver: () => ModelOptions
Useful to change the options used by the model of a task at runtime.
Tasks output events:
Example usage in the application task graph:
on('inputDataForHumanActivity', run('humanActivityClassificationUsingCNN')); on('humanActivityPredicted', run('writeRecords')); on('inputDataStressLevel', run('stressLevelRegression')); on('stressLevelPredicted', run('writeRecords'));
Note: To use the
writeRecords
task, the persistence package must be installed and configured. See persistence package docs.
Events
| Name | Payload | Description |
|--------------------------------|-------------------------------------|-----------------------------------------------------|
| {classificationAim}Predicted
| Classification
| Indicates that a classification has been completed. |
| {regressionAim}Predicted
| Regression
| Indicates that a regression has been completed. |
Records
Classification
| Property | Type | Description |
|------------------------|-------------------------------------------------|--------------------------------------------------------------------|
| id
| string
| Record's unique id. |
| type
| string
| Always {classificationAim}-prediction
. |
| change
| Change
| Always NONE
. |
| timestamp
| Date
| The local time when the model predicted the classification result. |
| classificationResult
| ClassificationResult
| Object containing the results of the classification. |
ClassificationResult
| Property | Type | Description |
|----------------|-----------------------------------------------------------|-----------------------------------------------------------------|
| prediction
| ClassificationPrediction[]
| Array of the classification predictions generated by the model. |
| modelName
| string
| The name of the model used for the classification. |
| architecture
| string
| The architecture of the model used for the classification. |
| version
| string
| The version of the model used for the classification. |
ClassificationPrediction
| Property | Type | Description |
|----------|----------|--------------------------|
| label
| string
| Identifier of the class. |
| score
| number
| Score of the prediction. |
Regression
| Property | Type | Description |
|--------------------|-----------------------------------------|---------------------------------------------------------------|
| id
| string
| Record's unique id. |
| type
| string
| Always {regressionAim}-prediction
. |
| change
| Change
| Always NONE
. |
| timestamp
| Date
| The local time when the model predicted the regression result |
| regressionResult
| RegressionResult
| Object containing the results of the regression. |
RegressionResult
| Property | Type | Description |
|----------------|------------|-----------------------------------------------------------|
| prediction
| number[]
| Array of numbers with the results generated by the model. |
| modelName
| string
| The name of the model used for the regression. |
| architecture
| string
| The architecture of the model used for the regression. |
| version
| string
| The version of the model used for the regression. |
License
Apache License Version 2.0
Disclaimer
While we state that CNN models are supported, only 1D-CNN models have been tested. The code is general enough to support 2D and 3D CNN models with one input tensor, but they have not been tested. If you try 2D/3D-CNN models and something is not working as expected, contact us.