model image_generation template risk: low
Subway Platform Candid Woman Portrait
Configures AI image generation for a photorealistic street-style portrait of an adult woman matching a reference identity, standing candidly on a subway platform. Specifies details…
PROMPT
{
"category": "SUBWAY_PLATFORM_STREET_CANDID",
"identity_lock": {
"enabled": true,
"priority": "ABSOLUTE_MAX",
"instruction": "Use reference image identity exactly. Adult 21+. Preserve face proportions and marks. No beautification."
},
"subject": {
"demographics": "Adult woman, 21-29, match reference identity.",
"hair": {
"color": "Match reference.",
"style": "Low ponytail or loose waves tucked behind scarf",
"texture": "Real strands; slight frizz; flyaways",
"movement": "Minimal movement, platform breeze subtle"
},
"face": {
"eyes": "Exact reference; reflective catchlights",
"skin_details": "Pores visible, realistic shadows",
"micro_details": "Preserve marks"
},
"clothing": {
"outerwear": "Minimal black coat or jacket (no logos/text)",
"extras": "Scarf optional (no patterns with text)",
"fabric": "Wool texture visible"
},
"accessories": {
"jewelry": ["Small silver hoops (optional)"],
"bag": "Simple tote/shoulder bag (no logos)"
}
},
"pose": {
"type": "Candid waiting",
"orientation": "Half-body standing near platform edge (safe distance)",
"head_position": "Slight tilt, calm posture",
"hands": "One hand holding bag strap, other in pocket",
"gaze": "Looking toward camera with neutral confidence",
"expression": "Calm, slightly serious"
},
"setting": {
"environment": "Subway platform",
"background_elements": [
"Overhead fluorescent lights",
"Train blur in background (no readable signage)",
"Platform tiles with realistic wear"
],
"depth": "Face sharp; background softened"
},
"camera": {
"shot_type": "Street-style portrait",
"angle": "Eye level",
"focal_length_equivalent": "35mm editorial OR 26mm phone",
"framing": "4:5, leading lines from platform",
"focus": "Eyes sharp, background motion blur allowed"
},
"lighting": {
"source": "Fluorescent overhead + ambient",
"direction": "Top-down with mild fill",
"highlights": "Realistic shine on hair/skin",
"shadows": "Soft, slightly cool subway contrast"
},
"mood_and_expression": {
"tone": "Moody, urban, confident",
"atmosphere": "Real city commute candid"
},
"style_and_realism": {
"style": "Photoreal street portrait",
"imperfections": "Noise + slight motion blur in background"
},
"technical_details": {
"aspect_ratio": "4:5",
"resolution": "High",
"noise": "Moderate low-light grain",
"mode_variants": {
"amateur": "Phone-like HDR, mild grain, imperfect framing",
"pro": "Cleaner exposure, controlled highlights, crisp subject separation"
}
},
"constraints": {
"adult_only": true,
"single_subject_only": true,
"no_text": true,
"no_logos": true,
"no_watermarks": true,
"no_readable_signage": true
},
"negative_prompt": [
"readable signs", "logos", "watermark",
"identity drift", "face morphing",
"extra fingers", "warped hands",
"cgi", "plastic skin", "over-smoothing"
]
} REQUIRED CONTEXT
- reference image
OPTIONAL CONTEXT
- mode_variants
EXPECTED OUTPUT
- Format
- image_prompt
- Constraints
-
- photorealistic
- match reference identity exactly
- include negative prompt
- aspect ratio 4:5
CAVEATS
- Dependencies
-
- Requires reference image.
- Missing context
-
- Reference image (URL, embedding, or description)
- Target AI image generation tool/model (e.g., Stable Diffusion version)
- Ambiguities
-
- Assumes a reference image exists but does not specify how it is provided or integrated.
- Hair style references 'tucked behind scarf' but scarf is optional.
QUALITY
- OVERALL
- 0.91
- CLARITY
- 0.92
- SPECIFICITY
- 0.95
- REUSABILITY
- 0.85
- COMPLETENESS
- 0.92
IMPROVEMENT SUGGESTIONS
- Add a 'reference_image' field with URL or instructions for IP-Adapter/ControlNet usage.
- Resolve hair-scarf dependency by providing alternatives: e.g., 'Low ponytail or loose waves (tucked behind scarf if present)'
- Include example full prompt string for direct copy-paste use.
USAGE
Copy the prompt above and paste it into your AI of choice — Claude, ChatGPT, Gemini, or anywhere else you're working. Replace any placeholder sections with your own context, then ask for the output.
MORE FOR MODEL
- Cinematic Image-to-Video Prompt Generatormodelimage_generation
- Product Image Studio Enhancement Transformermodelimage_generation
- Rooftop Golden Hour Bikini Portrait Generatormodelimage_generation
- Cheerful Student Home Study Scene Promptmodelimage_generation
- Corporate Logo Designer with Brand Variablesmodelimage_generation
- Bald Eagle Head Volumetric 3D Render Promptmodelimage_generation
- 360-Degree Product Rotation Video Generatormodelimage_generation
- Clothing Flat Lay to Ghost Mannequin Transformermodelimage_generation
- Cold War Spy Bridge Handover Image Editormodelimage_generation
- High-End Cinematic Realism Image Promptmodelimage_generation