Does it include support for 3D body pose?

Yes, it provides documentation for VNDetectHumanBodyPose3DRequest, which returns skeletal data in real-world meters.

What Apple platforms are supported by this reference?

This skill covers iOS, iPadOS, macOS, tvOS, and visionOS, including the latest features introduced in iOS 17 and macOS 14.

Does this skill help with text and barcode recognition?

Yes, it includes full API references for VNRecognizeTextRequest (OCR), VNDetectBarcodesRequest, and DataScannerViewController.

How does this differ from the 'vision' skill?

While the 'vision' skill focuses on high-level decision trees and patterns, 'vision-ref' provides deep-dive technical API signatures and specific landmark definitions.

Can I use this for subject lifting like in the Photos app?

Absolutely. It includes references for VNGenerateForegroundInstanceMaskRequest and VisionKit's ImageAnalysisInteraction.

Vision Framework API Reference

Name: Vision Framework API Reference
Author: CharlesWiltgen

byCharlesWiltgen

•

148

Mobile Development

Provides a comprehensive technical reference for Apple's Vision framework to implement advanced computer vision and pose detection.

关于

The vision-ref skill is an essential technical companion for xOS developers implementing computer vision capabilities. It provides instant access to detailed API signatures, landmark mapping, and implementation patterns for Apple's Vision and VisionKit frameworks. From subject lifting and instance segmentation to complex 2D/3D human body and hand pose detection, this skill ensures developers follow Apple's best practices for coordinate mapping and performance optimization. It is particularly useful for building gesture-driven interfaces, augmented reality effects, and sophisticated image analysis tools across iOS, macOS, and visionOS.

主要功能

Implementation guides for subject lifting and foreground instance segmentation.
Detailed reference for VNRecognizeTextRequest (OCR) and barcode detection.
Documentation for 3D skeleton detection and real-world coordinate systems.
Comprehensive mapping for 21 hand landmarks and 18 body landmarks.
148 GitHub stars
Performance best practices for background queue handling and resource management.

使用场景

Developing fitness or motion-tracking apps using body pose estimation.
Implementing 'Remove Background' or subject extraction features in creative apps.
Building gesture-controlled interfaces for hands-free app interaction.

关于

主要功能

Implementation guides for subject lifting and foreground instance segmentation.
Detailed reference for VNRecognizeTextRequest (OCR) and barcode detection.
Documentation for 3D skeleton detection and real-world coordinate systems.
Comprehensive mapping for 21 hand landmarks and 18 body landmarks.
148 GitHub stars
Performance best practices for background queue handling and resource management.

使用场景

Developing fitness or motion-tracking apps using body pose estimation.
Implementing 'Remove Background' or subject extraction features in creative apps.
Building gesture-controlled interfaces for hands-free app interaction.