Back to feed
arXiv cs.LG
arXiv cs.LG
7/21/2026
Fully-sensorized smart-eyewear platform for on-device Machine Learning

Fully-sensorized smart-eyewear platform for on-device Machine Learning

Short summary

ARGO is a smart eyewear platform that runs on-device ML using an STM32N6 microcontroller with integrated NPU, enabling real-time urban obstacle recognition without cloud dependency. It introduces Head-wise Parallel Attention to adapt YOLOv11 for the NPU, achieving 10 FPS with a 2.483 MB footprint and ~113 minutes battery life. The work demonstrates that competitive edge AI requires tightly integrated hardware, firmware, and model co-design.

  • •ARGO smart eyewear runs YOLOv11 on an STM32N6 NPU for on-device obstacle detection
  • •Head-wise Parallel Attention adapts the model for efficient NPU execution under 2.5 MB memory
  • •Platform delivers 10 FPS with ~113 min autonomy on a 200 mAh battery

Generated with AI, which can make mistakes.

#research-breakthrough#ai-tools#open-source
Read full article at arXiv cs.LG

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

I Built AI Smart Glasses That Respond in Under 2 Seconds — Here's How
Dev.to
Beyond the Cloud: Engineering "Micro-AI" on Consumer Hardware
Dev.to
The original title is "Mastering Edge AI: How to Build High-Speed Vision Analyzers on Android"
Dev.to
The original title is "The 16.67ms Race: Mastering Real-Time 60 FPS Video Segmentation on Android"
Dev.to