skills.rest
SharpAI
SharpAIOFFICIAL
Explore all skills by this author →
GitHub

yolo-detection-2026-coral-tpu-win-wsl

Detects objects in camera frames on Google Coral Edge TPU via Windows WSL.

3.0k|478|Updated Mar 5, 2019
Search Tags:#object detection#surveillance#yolo#wsL#TFLite#Real-time#coral edge tpu
One-click install
npx skills add https://github.com/SharpAI/DeepCamera --skill yolo-detection-2026-coral-tpu-win-wsl
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: yolo-detection-2026-coral-tpu-win-wsl
Source: https://github.com/SharpAI/DeepCamera/tree/main/skills/detection/yolo-detection-2026-coral-tpu-win-wsl
Command: npx skills add https://github.com/SharpAI/DeepCamera --skill yolo-detection-2026-coral-tpu-win-wsl

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires ai-edge-litert, numpy, Pillow, and includes scripts (resource) and assets (resource) components.

What problem does it solve?

Running real-time object detection on Windows with a Google Coral Edge TPU is difficult because the TPU drivers and runtime are Linux-native. This skill bridges that gap by executing detection inside WSL with automatic USB passthrough, delivering ~4ms inference without cloud dependencies.

Core Features & Use Cases

  • Hardware-Accelerated Detection: Runs YOLO 2026 nano compiled for Edge TPU, detecting 80 COCO classes (person, car, dog, cat) at up to 15 FPS.
  • Automatic USB Bridging: Uses usbipd-win to auto-attach the Coral USB Accelerator (hardware ID 18d1:9302) into WSL on startup.
  • TPU Health Watchdog: Detects inference hangs (10s timeout) and silent stalls (30 consecutive empty frames), emitting telemetry so the host can restart the process.
  • Use Case: A home security setup on a Windows PC uses a Coral USB Accelerator to detect people and cars on camera feeds locally, with frames exchanged via a shared /tmp workspace and results streamed as JSONL events.

Quick Start

Ask the AI to run real-time object detection on your camera feed using the Coral Edge TPU through WSL with default person, car, dog, and cat classes.

Frequently Asked Questions about yolo-detection-2026-coral-tpu-win-wsl

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I use a Coral Edge TPU on Windows for object detection?▼

Run the skill's deploy.bat, which verifies usbipd, binds the Edge TPU hardware IDs, and creates a Python virtual environment inside WSL. Detection then runs in WSL via detect.py while the Windows host exchanges frames and results over JSONL stdin/stdout.

What is the inference speed of Coral Edge TPU with YOLO models?▼

At 320x320 input the model maps fully onto the Edge TPU and runs in about 4ms per frame. At 640x640 some layers fall back to the host CPU, raising inference time to roughly 20ms.

Does Coral Edge TPU work inside WSL2 on Windows?▼

Yes, but the USB device must be passed into WSL using usbipd-win. The skill's wrapper script auto-attaches hardware ID 18d1:9302 to WSL on startup so the TPU is accessible to the Linux runtime.

Why does Edge TPU detection fall back to CPU inference?▼

Fallback occurs when the libedgetpu delegate fails to load, typically because the USB device is not attached or the driver is missing. The skill then loads a CPU-compatible tflite model such as SSD MobileNet instead.

What happens if the Coral TPU overheats or stalls?▼

A built-in watchdog hard-times-out any invoke call exceeding 10 seconds and flags a stall after 30 consecutive empty frames following prior detections. The host receives an error or warning event and can restart the process.

Related Skills

Explore More Software Engineering Skills→
gevreygevrey

ezviz-h1c-local

Validate EZVIZ H1c camera RTSP streams and capture proof frames on the LAN.

Community
Intermediate
LuminaVaultLuminaVault

Market Alert Service

Monitor stocks, crypto, and news markets and send alerts to Discord, Telegram, or Slack.

Official
Advanced
NVIDIA-TAONVIDIA-TAO
83

tao-generate-referring-expressions

Generate region descriptions, scene captions, and referring expressions from images and KITTI labels.

Official
Advanced
SharpAISharpAI
3.0k

yolo-detection-2026

Detect objects in video frames using YOLO 2026 models.

Official
Advanced