Smart Assistants for Board Games

BGLIB is an Android library for the development of smart assistants for board games (board games library). It utilises various computer vision techniques and machine learning algorithms. Packages like OpenCV, ML Kit and Mediapipe are already integrated.

Minimum SDK version: 30

Installation

To integrate a library into a project a user needs to add the jitpack repository into settings.gradle.kts as follows:

dependencyResolutionManagement {
    repositoriesMode.set(RepositoriesMode.FAIL_ON_PROJECT_REPOS)
    repositories {
        mavenCentral()
        maven { url = uri("https://jitpack.io") }
    }
}

Afterwards it is possible to integrate the library inside build.gradle.kts:

dependencies {
    implementation(libs.junacik99.bglib)
}

Or refer to https://jitpack.io/#Junacik99/bglib for detailed instructions on how to integrate the bglib library into an Android project.

Documentation

The docs are in the Documentation file.

Source Code

The source code of the library is in the bglib directory.

Used models are in the assets directory.

Train Your Own Model

To train a custom model for the image classification, you can use a model training pipeline in Airflow.

To install Airflow, follow these instructions.