DroneDetection (v1) — ⚠️ Deprecated

This model is superseded by DroneDetectionV.2.
Kept for archival reference only.

Part of OmniSense / MediaSense — Connectivia Labs' industrial AI platform.

Performance

Metric	Value
mAP@0.5	98.9%
Architecture	YOLO (v1 training run)
Input size	640×640

Classes

drone · uav

Training Notes

• First training iteration for drone/UAV detection
• Superseded by v2 with improved dataset and augmentation pipeline
• Hardware: NVIDIA GPU (GCP)

Status

🔒 Private — archived. Use DroneDetectionV.2 for all active deployments.

Platform

OmniSense — Multi-agent industrial AI platform
Pillar: MediaSense (Computer Vision)
Connectivia Labs — Internal use only