A 96pJ/Frame/Pixel and 61pJ/Event Anti-UAV System with Hybrid Object Tracking Modes

2025-12-12

AI Summary

To address the challenge of balancing energy efficiency and reliability in detecting small, high-speed unmanned aerial vehicles (UAVs), this paper proposes a low-power hybrid-sensing anti-UAV system. Methodologically, it introduces a novel frame-based and event-driven dual-mode adaptive tracking architecture that dynamically switches between modes based on target size and velocity; designs a trajectory-classification-guided zero-skipping MAC neural computing unit, an adaptive-threshold tracking module, and a run-length encoding–based event reconstruction mechanism. Key contributions include: >97% reduction in neural computation redundancy and state-of-the-art end-to-end energy efficiency; a 2 mm² chip fabricated in 40 nm CMOS achieving ultra-low power consumption of 96 pJ/frame/pixel and 61 pJ/event; and 98.2% detection accuracy across operational ranges of 50–400 m and target velocities of 5–80 px/s.

Technology Category

• Machine Learning: Hardware-aware ML
• Computer Vision: Motion & Tracking
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Security and Privacy: Tracking, profiling, and countermeasures against them
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We present an energy-efficient anti-UAV system that integrates frame-based and event-driven object tracking to enable reliable detection of small and fast-moving drones. The system reconstructs binary event frames using run-length encoding, generates region proposals, and adaptively switches between frame mode and event mode based on object size and velocity. A Fast Object Tracking Unit improves robustness for high-speed targets through adaptive thresholding and trajectory-based classification. The neural processing unit supports both grayscale-patch and trajectory inference with a custom instruction set and a zero-skipping MAC architecture, reducing redundant neural computations by more than 97 percent. Implemented in 40 nm CMOS technology, the 2 mm^2 chip achieves 96 pJ per frame per pixel and 61 pJ per event at 0.8 V, and reaches 98.2 percent recognition accuracy on public UAV datasets across 50 to 400 m ranges and 5 to 80 pixels per second speeds. The results demonstrate state-of-the-art end-to-end energy efficiency for anti-UAV systems.

Problem

• Develops an energy-efficient anti-UAV system for detecting small, fast drones
• Integrates frame-based and event-driven tracking to improve reliability
• Reduces redundant neural computations for enhanced energy efficiency

Innovation

• Integrates frame-based and event-driven object tracking
• Uses adaptive switching between frame and event modes
• Implements zero-skipping MAC architecture for efficiency