Alpaca help people with dyslexia to read a text from a paper using OCR to generate a text-to-speech service and user's preference customizable text style. Alpaca is part of Bangkit Academy 2022 capstone project from C22-PS068.
| Bangkit ID | Name | Learning Path |
|---|---|---|
| M2002F0103 | Shaffira Alya Mevia | Machine Learning |
| M7224W2084 | Hanifah Putri Alamsyah | Machine Learning |
| M2010G1081 | Nisrina Dinda Dhamayanti | Machine Learning |
| A7200F1868 | Fikri Yusrihan | Android Development |
| A2200F1869 | Muhammad Fatkhu Rohman | Android Development |
| C2009G0971 | Jericho Gunawan | Cloud Computing |
- Make datasets or find public dataset for Optical Character Recognition
- Find library to easier when making model Machine Learning
- For OCR dataset, we use TRDG dataset from (https://bit.ly/ALPACADataset)
- For Text-to-Speech, we use gTTS library
- Make two models for Susun Kata game (TTS) and photo scanning (OCR) features
- Save and export the existing model to .tflite file format
- Create Backend using JavaScript Programming Language and Express.js
- Creating API for access data
- integrate with Text-to-Speech services provided by Google Cloud Platform
- Deployment on Google Compute Engine
- Create end-to-end application design, start from a wireframe until a high-fidelity prototype
- Implement an Android-based application using Kotlin programming language
- Implement MVVM Architecture into the application
- Integrate with the Machine Learning model using TFLite library
- Integrate with the backend services provided by Google Cloud Service using Retrofit and Firebase
The deployment save in (https://bit.ly/ALPACADeploymentLink)
Reference: [1] TRDG DATASET [2] Plate Character Recognition [3] Train TRDG Dataset