New system enables real-time aerial person tracking on low-power hardware

Researchers have developed a novel system called EMTS-Det for real-time aerial person tracking on low-power hardware. This five-stage system estimates ego-motion, normalizes frame data, and uses a small, efficient network to detect and track individuals. The system achieves significantly higher frame rates and accuracy on resource-constrained devices like the Raspberry Pi Zero 2W compared to standard models like YOLOv8n, demonstrating its effectiveness in challenging real-world UAV video scenarios. AI

IMPACT Enables real-time AI-powered tracking on low-power drones, potentially for surveillance or autonomous navigation.

RANK_REASON Academic paper detailing a new computer vision system and its performance evaluation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• EMTS-Det
• Gholamreza Anbarjafari
• Raspberry Pi Zero 2W
• VisDrone-DET
• YOLOv8n

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →