skills.rest
limit5
limit5COMMUNITY
Explore all skills by this author →
GitHub

npu-pose

Deploy pose estimation models to NPU hardware with INT8/FP16 quantization tuning.

1|Updated Apr 9, 2026
Search Tags:#mediapipe#pose estimation#npu deployment#keypoint verification#oks metric#movenet#skeleton tracking
One-click install
npx skills add https://github.com/limit5/OmniSight-Productizer --skill npu-pose
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: npu-pose
Source: https://github.com/limit5/OmniSight-Productizer/tree/main/configs/skills/npu-pose
Command: npx skills add https://github.com/limit5/OmniSight-Productizer --skill npu-pose

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Deploying human pose and gesture estimation models to NPU hardware often leads to reduced keypoint accuracy due to quantization, making it difficult to meet performance requirements for real-time edge applications like AI camera body tracking.

Core Features & Use Cases

  • Model Selection Guidance: Choose between top-down and bottom-up pose estimation architectures, with recommendations for MoveNet, HRNet, and MediaPipe Pose models.
  • NPU Quantization Tuning: Optimize model performance with INT8 quantization, with automatic fallback to FP16 if Object Keypoint Similarity (OKS) scores drop below acceptable thresholds.
  • Accuracy Validation: Verify deployment performance using COCO keypoint validation datasets, with per-joint accuracy analysis and temporal jitter checks for video sequences.
  • Use Case: Ideal for embedded AI camera systems that require low-latency, accurate body tracking for applications like retail analytics, security monitoring, and fitness tracking.

Quick Start

Use the npu-pose skill to deploy your selected pose estimation model to the target NPU and confirm it meets the required [email protected] accuracy and latency thresholds.

Frequently Asked Questions about npu-pose

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

FAQPage Schema
How do I deploy pose estimation models to NPU hardware without losing keypoint accuracy?▼

To deploy pose estimation models to NPU hardware without losing keypoint accuracy, apply INT8 quantization with automatic FP16 fallback if OKS scores drop below acceptable thresholds.

What's the best way to reduce temporal jitter in video-based pose inference on edge devices?▼

Reducing temporal jitter in video-based pose inference requires applying temporal jitter checks during the accuracy validation phase after deploying your selected pose estimation model to the NPU.

How do I validate keypoint accuracy after quantizing a pose estimation model for edge computing?▼

To validate keypoint accuracy after quantization, verify deployment performance using the COCO keypoint validation datasets, conducting per-joint accuracy analysis and temporal jitter checks.

Can I use MediaPipe or MoveNet models for embedded AI camera body tracking on NPUs?▼

Yes, you can use MediaPipe Pose, MoveNet, or HRNet models for embedded AI camera body tracking on NPUs, with guidance provided for choosing between top-down and bottom-up architectures.

Why does INT8 quantization drop OKS scores for skeleton tracking on edge devices?▼

INT8 quantization drops OKS scores for skeleton tracking because reducing model precision degrades keypoint detection accuracy, requiring FP16 fallback to maintain acceptable thresholds.

Do I need COCO keypoint datasets to verify NPU pose model deployment?▼

Yes, COCO keypoint validation datasets are needed to verify NPU pose model deployment, enabling per-joint accuracy analysis and temporal jitter checks for video sequences.

Related Skills

Explore More Software Engineering Skills→
ThedouglerThedougler

mediapipe-pose-detection

Debug MediaPipe pose detection by adjusting confidence thresholds and validating landmarks.

Community
Intermediate
limit5limit5

npu-barcode

Detect and decode barcodes and QR codes on NPU-enabled embedded cameras.

Community
Advanced
limit5limit5

npu-detection

Deploy YOLO models to NPU hardware with quantization and accuracy validation.

Community
Intermediate
melissa-pereira-deelmelissa-pereira-deel

On-Device ML Optimization

Optimize on-device ML models for speed, size, and energy efficiency.

Community
Advanced