vision▌

図表が多い PDF を画像化して、Claude の vision 機能で内容を解析・Markdown 化するスキルです。

Comprehensive guide to using Salesforce's BLIP-2 for vision-language tasks with frozen image encoders and large language models.

Detect text, faces, barcodes, objects, and body poses in images and video using on-device computer vision. Patterns target iOS 26+ with Swift 6.3, backward-compatible where noted.

Apple Vision Framework for computer vision tasks: subject segmentation, pose detection, text recognition, barcode scanning, and document processing. \n \n Covers 13+ Vision APIs across subject lifting, hand/body pose, person segmentation, text OCR, barcode detection, and document scanning with decision trees for choosing the right tool \n Includes 15 production patterns: combining APIs to exclude hands from objects, real-time gesture recognition, multi-person segmentation, fitness action classif

axiom-vision-ref