Pickai-pose-gallery-zh: 1439 AI Pose Data Points, Translated from Japanese to Chinese

A Chinese-language gallery that systematically organizes human pose references needed for AI drawing and 3D work. It provides a full translation and mirror of 1439 poses and 2878 images based on the original Japanese data. Each pose consists of high-resolution PNG originals and low-capacity JPG thumbnails captured from 5 angles (front, side, top, bottom, etc.).

1 / 2

A key feature is its zero-dependency structure, running on pure HTML and JSON without frameworks or external libraries. Filtering and search are available based on 4 main categories (full body, upper body, two-person, etc.) and 26 detailed action classifications. It supports keyboard shortcuts and mobile gestures, allowing efficient image browsing in both desktop and mobile environments.

It provides a complete static site template that developers can host or modify themselves, rather than just a simple image collection. It includes a Cloudflare Pages deployment script and data validation tools, making it immediately deployable for personal portfolios or learning purposes. The original data and translated JSON files are separated, making it easy to analyze the data structure or extend custom search functionality.

It is useful for AI art creators or frontend developers to reference how to build static sites and handle large-scale image processing. It contains adult content, so care must be taken regarding the purpose and audience of use. Image copyright belongs to the original creators, and this repository is recommended for use only for learning and archiving purposes regarding code and data structure.