AI & Computer Vision / 2025
DisasterLens
A real-time, decentralized disaster detection system built with YOLOv8 and edge computing to detect environmental anomalies while optimizing AI models for resource-constrained environments.
- Role
- Core Engineer & Researcher
- Stack
- PythonYOLOv8Edge AIComputer VisionMachine Learning
Problem
Disaster detection systems often depend on centralized infrastructure and computationally expensive models, creating challenges for real-time detection in resource-constrained environments.
Process
Conceptualized and developed a decentralized disaster detection approach using computer vision and edge computing, focusing on real-time environmental anomaly detection and model optimization.
Solution
Built a disaster detection system utilizing YOLOv8 and edge computing to enable real-time detection while optimizing the model for constrained computational environments.
Result
Successfully pitched the project through the Find IT! UGM 2025 Hackathon and secured a Top 15 position.
Gallery