Loading…
Loading…


Every Piece Begins with a Feeling — Shaped by Hands, Code & AI
每件作品都從一個念頭開始,用手、用程式、用 AI 把它做出來
I'm a physician in Taiwan who can't stop making things — cold porcelain flowers, hand-crafted paper art, AI-generated films and music, and small web tools built with code. This site collects the works and the stories behind them.


精選作品
AI FilmAn AI-generated short film confronting antimicrobial resistance and clinical fear, told through the eyes of an infectious disease physician.
AI FilmA steampunk waltz music video set in a fogbound city, where the Cat Duke sings a gentle nocturne for streetlamps and drifting mist.
handcraftedUltra-realistic botanical sculptures in cold porcelain clay — ranunculus, peonies, foxgloves, and lily of the valley.
創作工具與實驗
ToolAn AI-powered creative operating system for stickers, images, layout, and print workflows. 19 professional tools from AI generation to print-ready output.
ToolCompose motifs into seamless tiles, inspect with seam heatmap, and export print-ready PNG/PDF. 10 repeat modes with print size calculator.
ToolQuickly try on, recolor, and generate outfits with AI. A fashion experiment platform combining generative AI with virtual try-on technology.
創作幕後、工具分享與靈感隨筆
AI Image17 組動漫 prompt、同一組 seed,實測 Anima 全家四個版本(含社群加深版 2.9B)、SDXL 動漫王者 NoobAI-XL、新架構 NetaYume Lumina、Z-Image-Anime,再拉泛用模型 Krea2 當對照組。打鬥誇飾、手部完整度、風格多樣性一次測完——動漫題到底該不該用動漫專精模型?
ComfyUI把 ComfyUI 裡找得到的放大模型與臉部修復模型全部接進同一張工作流,用 12 張有「標準答案」的測試圖(先生成高清原圖、再程式降質成 128/256/512 像素與仿老照片)加 3 張真老照片,同圖實測 37 條路線,除了看圖還算了「像不像原人」的分數。哪款該當日常主力、修臉為什麼會換臉、老照片該找誰、硬碟裡那一堆放大模型哪幾款值得留,這篇從放大原理講起,一次說清楚。
ComfyUI把 ComfyUI 裡找得到的去背模型全部接進同一張工作流,用 15 張專門找碴的圖(髮絲、玻璃、煙、頭紗、Logo、動漫、單色背景)同圖實測 30 條路線。哪一款該當日常主力、半透明要找誰、先把背景改成反差色再去背到底好不好用,這篇從去背原理講起,一次說清楚。



A doctor who can't stop making things — from hand-sculpted flowers to AI-generated films, each piece starts with a feeling that refused to stay in my head.
一個停不下來的醫師創作者——從手捏的花到 AI 生成的影片,每件作品都是一個不肯只留在腦海裡的念頭。