GreenEye: AI-Powered Indonesian Plant Recognition

GreenEye is an AI-powered plant recognition application designed to help users identify indigenous Indonesian flora through real-time image classification. By combining Computer Vision with a mobile-first user experience, the application enables users to recognize plants such as pineapple (nanas), banana (pisang), and other local species simply by using their smartphone camera.

The project aims to make botanical knowledge more accessible to the public while demonstrating the practical application of deep learning in mobile-integrated environments.

🛠️ My Role: AI Engineer

As the AI Engineer for this project, I was responsible for developing and optimizing the image classification pipeline used for plant recognition.

Key Responsibilities

• Model Development & Fine-Tuning
  Fine-tuned the MobileNetV2 architecture on a curated dataset of Indonesian plant species to achieve an efficient balance between classification accuracy and inference performance.

• Cloud-Based Inference Deployment
  Prepared and optimized the trained model for server-side deployment, enabling real-time predictions for mobile users through cloud-hosted inference.

• Dataset Curation & Preprocessing
  Collected, cleaned, and preprocessed plant imagery captured under varying environmental and lighting conditions to improve model robustness and generalization.

• Inference Workflow Integration
  Contributed to the integration workflow between the mobile application and the AI inference service, ensuring responsive image classification results.

🚀 Key Features

• Real-Time Plant Identification
  Users can capture or upload plant images and receive instant classification results powered by a deep learning model.

• Educational Plant Information
  The application provides additional information about identified plant species to promote awareness of Indonesian biodiversity.

• Mobile-Centered User Experience
  Designed to deliver a simple and accessible user experience for non-technical users interested in learning about local flora.

🎯 Technical Takeaways

Working on GreenEye strengthened my understanding of deploying Computer Vision models in real-world applications. Beyond training accurate models, I learned the importance of designing efficient inference pipelines, handling diverse image data, and balancing prediction responsiveness with deployment constraints in cloud-connected mobile systems.

The project also gave me practical experience in applying deep learning to production-oriented workflows, particularly in the context of mobile-integrated AI services.