時尚廣告牌產生器
將服裝照片轉為專業時尚模特的商場廣告。
關於
The Fashion Billboard Generator workflow is designed to transform ordinary clothing photos into stunning, professional mall billboard advertisements. This workflow leverages the power of AI to seamlessly integrate clothing items onto realistic fashion models, creating eye-catching visuals suitable for high-impact advertising. At the core of this process is the OpenAIGPTImage1 node, which utilizes advanced generative models to render high-quality images that blend the clothing items with model figures.
The workflow begins with the LoadImage node, which imports the original clothing photo. This image is then processed through the OpenAIGPTImage1 node, where the AI model performs the transformation, ensuring that the clothing fits naturally on the model's body. Finally, the SaveImage node stores the completed billboard-ready image for easy access and distribution.
您提供的內容
- 正在讀取工作流程
在 ComfyUI 中執行。 此工作流程無法在此頁面上執行。請在已載入所有內容的 Comfy Cloud 中開啟,或下載到您自己的 ComfyUI。
在 Cloud 中試用工作流程內部
節點圖的唯讀預覽。請在 Comfy Cloud 中開啟,或下載後編輯節點。
Loading the graph
Open full-size workflow preview- 作者
- ComfyUI
需要自訂節點
- 正在讀取工作流程
從程式碼使用此工作流程
可行時會根據工作流程圖和輸入內容產生;否則會顯示儲存的範例。執行可能產生費用的程式碼前,請檢查認證資訊和端點。
What you need
- Runs on
- Comfy Cloud
- 金鑰
- COMFY_API_KEY
- 您的檔案
- 由程式碼透過其 URL 上傳
需要付費 Cloud 方案與點數。
# Python 3.10+: pip install comfy-sdk==0.4.0
import os
from comfy_sdk import Comfy
api_key = os.environ.get("COMFY_API_KEY", "").strip()
if not api_key: raise RuntimeError("Set COMFY_API_KEY")
os.environ.setdefault("COMFY_BASE_URL", "https://cloud.comfy.org")
client = Comfy(api_key=api_key)
workflow = client.workflows.from_json({
"2": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "gpt-image-1.5",
"images": [
"10",
0
]
},
"_meta": {
"title": "Save Image"
}
},
"3": {
"class_type": "LoadImage",
"inputs": {
"image": "UPLOADED_image_FILENAME"
},
"_meta": {
"title": "Load Source Clothing"
}
},
"10": {
"class_type": "OpenAIGPTImage1",
"inputs": {
"background": "auto",
"custom_height": 1024,
"custom_width": 1024,
"image": [
"3",
0
],
"model": "gpt-image-1.5",
"n": 1,
"prompt": "dress a professional female model with the provided clothing items. Crucially, preserve ALL original material properties, textures, and fabric characteristics exactly as shown in the source clothing images.\n\nModel requirements:\n- Professional fashion model with natural, confident pose\n- Casual yet elegant stance suitable for contemporary streetwear\n- Natural facial expression, looking slightly away from camera\n- Modern hairstyle, effortless and stylish\n\nCritical fabric and material preservation:\n- Maintain EXACT fabric textures from the source images (fleece, denim, cotton, etc.)\n- Preserve the fuzzy, fluffy texture of the grey hoodie exactly as shown\n- Keep original fabric sheen, matte finish, or glossiness unchanged\n- Retain precise color values, saturation, and undertones of all garments\n- Preserve fabric weight appearance (heavy fleece vs. light cotton)\n- Maintain weave patterns, knit structures, and material densities\n- Keep all surface details: pile height, fabric grain, textile characteristics\n- Preserve stitching details, seam types, and construction elements\n\nClothing fit and draping:\n- Fit the provided clothing naturally to the model's body while maintaining fabric properties\n- Natural fabric behavior based on original material type (stiff denim vs. soft fleece)\n- Realistic folds, wrinkles, and creases appropriate to each fabric type\n- Proper layering with each garment's original thickness and volume\n- Accurate shadows and highlights respecting original material reflectance\n\nTechnical requirements:\n- Match lighting interaction with each fabric type from source images\n- Preserve fabric subsurface scattering (especially for fleece/fuzzy materials)\n- Maintain original color temperature and saturation of all garments\n- Keep exact contrast levels between different fabric areas\n- Professional fashion photography lighting showcasing material details\n- Clean neutral background (white or light grey)\n- High-resolution, editorial quality with sharp fabric texture detail\n- Photorealistic rendering with microscopic attention to textile authenticity\n\nEnvironment:\n- Lighting should reveal fabric textures clearly\n- Color-neutral lighting to preserve original garment colors\n- Studio-quality illumination showing material properties accurately\n\nDo not alter, enhance, or \"improve\" the fabric appearances. Transfer the exact material characteristics from the source clothing images to the dressed model. Preserve all design elements, logos, graphics, and details exactly as shown. No watermarks or additional text.",
"quality": "low",
"seed": 218519763,
"size": "1024x1536"
},
"_meta": {
"title": "OpenAI GPT Image 2"
}
},
"14": {
"class_type": "OpenAIGPTImage1",
"inputs": {
"background": "auto",
"custom_height": 1024,
"custom_width": 1024,
"image": [
"10",
0
],
"model": "gpt-image-1.5",
"n": 1,
"prompt": "Create a luxury fashion advertisement displayed on a large freestanding lightbox billboard inside a premium shopping mall.\n\nAdvertisement content (the image displayed ON the billboard):\n- The fashion model in a sophisticated contemporary art gallery setting\n- Background: minimalist white walls with abstract modern art pieces, floor-to-ceiling windows showing city skyline\n- Architectural elements: polished concrete floors, dramatic natural lighting from above\n- Model positioned dynamically in the space, interacting naturally with the environment\n- Subtle luxury details: marble accents, geometric sculptures, refined interior design\n- Professional editorial fashion photography with cinematic composition\n- Color palette: cool grays, warm accents, high contrast, premium aesthetic\n\nBillboard specifications:\n- Large-scale vertical/portrait format lightbox (approximately 3m x 5m)\n- Premium LED backlit display with even, bright illumination\n- Sleek aluminum or black matte frame with minimal bezels\n- Positioned prominently in mall central atrium or luxury brand corridor\n\nMall environment:\n- High-end shopping center with contemporary architecture\n- Polished marble or granite flooring reflecting the billboard light\n- Modern glass railings and multiple levels visible in background\n- Ambient mall lighting: warm and inviting but not overpowering the billboard\n- Subtle hints of luxury brand storefronts in soft focus background\n- Few elegantly dressed shoppers walking by (blurred, for context)\n- Potted architectural plants or modern sculptures nearby\n\nLighting and atmosphere:\n- Billboard as the focal point with professional illumination\n- Soft glow emanating from the lightbox\n- Upscale mall ambiance with layered architectural lighting\n- Natural daylight filtering through mall skylights\n- Reflections on polished floors adding depth\n- Warm, inviting yet sophisticated mood\n\nTechnical quality:\n- Shot from slight low angle to emphasize billboard presence and scale\n- Shallow depth of field keeping billboard in sharp focus\n- Photorealistic architectural photography\n- High dynamic range with rich details\n- Professional commercial photography standard\n- No text, watermarks, or additional branding\n\nThe overall scene should convey an aspirational luxury lifestyle brand in a premium retail environment, with the advertisement itself showcasing the clothing in an artistic, high-fashion context.",
"quality": "low",
"seed": 544862039,
"size": "auto"
},
"_meta": {
"title": "OpenAI GPT Image 2"
}
},
"16": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "gpt-image-1.5",
"images": [
"14",
0
]
},
"_meta": {
"title": "Save Image"
}
}
})
# Replace each example.com URL with a public URL of your file.
workflow.set_input("3", "image", client.assets.from_url("https://example.com/replace-with-a-url/image"))
# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)問答
可以,流程支援多種服飾照片,但建議使用高解析度、細節清晰的圖片以獲得最佳效果。
流程採用 OpenAIGPTImage1 模型,擅長將服飾自然融合於時尚模特兒身上。
OpenAIGPTImage1 節點會自動調整服飾貼合度,必要時可於節點內微調參數以提升效果。
流程本身無數量限制,但處理時間會依圖片大小與系統效能而異。