See our prototype here.
Nusatala.Prototype.mp4
- Nusatala
- Table of Content
- About Nusatala
Nusatala is a mobile application that can recognize images of various traditional musical instruments and provide relevant information. This application will have image scanning, recommendations for nearby art communities, articles, quizzes, faq, user profiles, and traditional musical instrument store features.
Nusatala aims to enhance our knowledge of Indonesian traditional music instruments by using an app that can recognize images of these instruments and provide relevant information. Although there are apps available for learning about these instruments, they only offer limited info and cannot identify images using machine learning. Using an app that can identify images and provide relevant information, Nusatala can easily enhance our knowledge about Indonesian traditional music instruments in one convenient app.
This step will explains how to install and deploy briefly :
- Clone this repository
git clone https://github.com/Nusatala/ML
- Download the dataset
Link dataset - Make a new virtual environment using Python
python -m venv nusatala-experimental
- Activate the virtual environment
env\Scripts\activate
- Install All the Requirements Inside "requirements.txt"
pip install -r requirements.txt
- Go to folder Google-Cloud-Deploy to activate Flask web server
cd Google-Cloud-Deploy - Run Flask
flask run
- Stop the application program or server by
ctrl + c.
Add files using the command line or push an existing Git repository with the following command:
cd existing_repo
git remote add origin https://gitlab.com/z4ed.thalib123/nusatala-api-docs.git
git branch -M main
git push -uf origin main- Nusatala app gets user image input with png, jpg, and jpeg format of traditional music instrument such as angklung, bonang, etc.
- Convert the image to Numpy then normalized to 255 scale
- Predict the image input using nusatala.h5 model that already build with ResNet and InceptionV3 pre-trained algorithm.
- The predicted output is name of the traditional music instrument. The result must following the label in dataset, that are "Bonang", "Kolintang", "Rebab", "Saluang", "Sape", "Sasando", and "Tifa"
- Finally, based on the traditional music instrument, system will render the regional origin, manufacture materials, history, tutorial to play the instrument video, and the community recommendation.
- Data input for Indonesian traditional music instrument classification is an image
- Image formatted as JPG within size 150x150 pixels
- Classification only for Indonesian traditional music instrument
- On the first development, only using several classification on traditional music instruments.
- The store is just for recommendation, we do not provide an e-commerce system for buying and sell traditional music instruments.
- Art groups recommendation for learning traditional music instruments are based on Google Maps
To keep on track, Nusatala team has been used Gantt Chart with Agile method. Additionally, Nusatala used to discuss for weekly sprint. Here is our Gantt Chart :
Our dataset can be found here Link dataset.
Tools/IDE/Library and resources that Nusatala use to build the app :
- Text Editor
-
- IDE
-
- Library
- Retrofit
-
- Platform
- Google Colaboratory
-
-
-
-
- API
- Google APIs
- Project Management
-
- [1] Hastawan, A.F. et al. (2019) ‘Designing Educational Game of Indonesian Traditional Musical Instruments Based on Android Using Unity 3D’, 379(Veic), pp. 92–100. Available at: https://doi.org/10.2991/assehr.k.191217.016.
Nusatala final presentation can be found here :
Bangkit Academy led by Google, GoTo, and Traveloka 2023 Batch 1
C23-PR552 Capstone Team - Nusatala
- [ML] M305DKX3952 - Guntur Aji Pratama
- [ML] M169DSX1306 - Mochammad Iqbal Syaifurrahman
- [ML] M340DSY3395- Reni Setyaningsih
- [CC] C040DSX2415 - Iqbal Alfarizi
- [CC] C066DSX0786 - Zaed Abdullah
- [MD] A066DSX1060 - Taufik Anwar