Projects
Waste sorting is a crucial method to reduce the amount of trash that ends up in landfills. It helps separate recyclables, organic waste, and non-recyclables, and prevents dangerous chemicals from contaminating other waste. The separation of the recycled trash will also help it to be recycled based on their respective methods. For our Artificial Intelligence course, we developed a live object detection AI for garbage classification (e.g., paper, metal cans, batteries) to help categorize waste in surroundings, landfills, or recycling facilities. My main contribution was developing the AI, from data preparation to model testing using a device camera. This project taught me the basics of machine learning (data preparation, splitting, training, and validation), the limitations of machine learning (particularly in detecting transparent objects like glass or plastic), and how to utilize and access my local camera through Visual Studio.
For my Human and Computer Interaction course, I completed an individual final project where I designed and developed a toy store website using HTML, CSS, JavaScript, and Figma. From this project, I gained a deeper understanding of Figma by utilizing its components, color schemes, design hierarchy, and design norms, navigation, and element placement. I also applied these design principles to create a coherent, functional, and aesthetically pleasing website. Additionally, I learned how to use HTML, CSS, and JavaScript to transfer my design into a responsive, working website.
ElderEase was the final group project for our Software Engineering course. We developed an application to ease communication between elderly individuals and their caretakers and to assist those with sensory impairments in navigating their surroundings. The app featured text-to-speech AI (to help those with declining eyesight), a daily word button (allowing elderly users to communicate needs with a single tap), an emergency call button, and reminders. We learned to develop applications following the Software Development Life Cycle, design user-friendly interfaces tailored to our target users (e.g., large fonts, color-coded buttons for those with vision problems), and integrate AI into an application.
Pages
- Login/Register
- Main
- Add New Activity
- Add Reminder
- Scan Text
Functionality
- Text-to-Speech
- Text Scanner
- Emergency Call
- Add New Activity
- Reminder
Collaborators
Sounify is an accent detection model. As people come from different regions, their native languages influence the way they speak, even when speaking other languages. This results in accents that can make understanding difficult during conversations. Recognizing a speaker’s accent is the first step toward improving comprehension. Sounify was a group project for our Speech Recognition course, where we used machine learning models such as Random Forest, KNN, and neural networks (CNN) to identify accents from audio. I primarily contributed to data preprocessing and training machine learning models (Random Forest, KNN, Decision Tree) in two scenarios: a 5-second audio dataset and a 10-second audio dataset. Through this project, we learned how to extract features from raw audio and apply preprocessing techniques.
Functionality
Accent Recognition
Collaborators
mTiks, inspired by m-Tix apps, was a personal Laravel project assigned on laboratory assistant training, where I was required to create a 10 page website using Laravel. The project criteria included applying the model, view, and controller layers (seeders were allowed), performing CRUD operations, implementing pagination, a search feature, Laravel validation, and middleware with user roles. As my first Laravel project, I used m-Tix as a design reference. Through this project, I learned about the application of design patterns in large websites and how to build a website using the Laravel framework.
In Indonesia, waste management is still poorly executed in some areas due to a lack of public concern, information, or proper management. Many people are unaware of the importance of waste management and the issues it causes, which leads to improper disposal methods such as burning, open dumping, or illegal burying, resulting in further pollution and contamination of the soil, air, and water. This project is a mobile app designed to address this issue by utilizing AI to assist with waste sorting and it is equipped with features such as waste reporting for unmanaged waste, city cleanliness statistics, points and vouchers to encourage proper waste management, and educational articles for users.
Pages
- Splash Screen
- Login/Register
- Statistical Menu
- Object Detection
- Article
- Report Menu
- Edit Profile
- History
- Vouchers
Functionality
- Navigation
- Components
- Animation
Collaborators
We used Naive Bayes, Neural Network, TFIDF Vectorizer, BERT, and data preprocessing technique to find the optimum news classification model with categories : entertainment, world, sports, techonolgy, politics, automobile, science
Functionality
Classify News into 7 Categories
Collaborators
DaResto is a restaurant simulation which runs on console. It has waiters, chefs, and can generate customers. It implements design patterns such as singleton, factory, observer, mediator, and facade. It uses multithreading where the customers, chefs, and waiters can run concurrently, changes through their states, and interact with each other.
Certificates
