- Project Description
- Built With
- App Prerequisites
- How Others Can Replicate
- How to Install App
- How to Use App
- Contributing
- Contact
- Others
- License
Front Page | Article Page | Detection | Result |
---|---|---|---|
Hi, introduce Dermate, an automatic skin diseases diagnostic application using image recognition. This application is specifically intended to detect skin diseases in babies. Dermate comes as a solution to solve problems and help the Indonesia government in the healthcare sector. As a healthcare application, we help people to treat diseases without having to go to the hospital directly during this pandemic COVID-19. Parents can simply capture image for their baby skin condition, and our smart AI can easily diagnostic their baby skin diseases. Our application is also supported by articles that are able to provide information to parents about the baby skin, and further treatment to overcome the disease. With Dermate we provide easy, fast, and inexpensive prevention of baby skin diagnosis with secure medical privacy.
- Android Studio Jetpack MVVM Architecture
- Google Cloud Platform Services GCP Console
ML model built with framework and libraries :
- Jupyter Notebook or Google Colab
- python==3.7.10
- tensorflow==2.5.0
- numpy==1.19.5
- scikit-learn==0.22.2
- matplotlib==3.2.2
- splitfolders==0.4.3
- keras==2.5.0
- MobileNet Model Based Architecture
- Android Version 9+
- Internet Connection
Built Model :
- Download our dataset from our drive
- Install all framework and libraries that needed
- Run our model on your colab
- Download model TFlite and labels.txt
Deployment Model :
- Initialize database
- insert data into the base for feature usage
- Initialize the intrepeter
- Preparing the image input (224x224x3)
- Perform inference
- Obtain map results
Mobile Application Development :
- compileSdkVersion 30
- buildToolsVersion "30.0.3"
- Download apk from the latest release
- Enable install from unknown sources
- Install apk
- Open app
- Tap the scan button in home page
- Add image
- Start recognize
- See result
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change. Next steps:
- Fork the project repo on GitHub
- Clone the project to your local machine
- Commit changes to your own branch
- Push your work back up to your fork
- Submit pull request, so we can review your changes
Note : Please make sure to update tests as appropriate.
- Putu Wahyu Tirta Guna - [email protected]
- Ilham Hadisyah Ramadhan - [email protected]
- Stenli Tong - [email protected]
- Putu Althea Putri Wiradani - [email protected]
- Ni Putu Yulika Trisna Wijayanti - [email protected]
- Ni Made Sinta Wahyuni - [email protected]
- Dermate Team - [email protected]
- Academic Resource Paper : MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- Project Link : Github Repository
- Presentation Document : [B21-CAP0022] Bangkit Capstone “Dermate” Automatic Skin Diseases Diagnostic Application using Image Recognition
Copyright (c) 2021 Sarjana Bangkit