Real-Time Obstacle Detection
Designed and implemented a real-time obstacle detection system using YOLO-based object detection models. Curated a custom dataset using CVAT annotation tools, trained and fine-tuned the model, and published both the dataset and trained models as open-source resources on Hugging Face and Kaggle. The pipeline runs at real-time speeds suitable for embedded deployment.
Stack & Tools
Links
Other Projects
Custom Quadcopter Build
End-to-end embedded build — PCB design in Altium, ArduPilot firmware, DJI flight controller integration, and full PID tuning.
ATF Robotics Cup
National championship-winning autonomous robot. Full system from PCB design and firmware to competition strategy.
FIRA RoboWorldCup
2nd place nationally and qualified for the World Finals. Embedded firmware and real-time control logic under competition conditions.