# gemini

Rotate the person in this image so they face directly forward toward the camera, with the head and shoulders square to the lens. Preserve the person's identity, eye shape, nose, jawline, hair length, hair color, hairstyle, eyebrow shape, lip color, makeup state, clothing color, clothing fit, accessories, and background exactly as the source. Adjust only the head, neck, and shoulder angle so both ears are visible, the gaze is straight at the camera, and the torso is centered. Output at the original resolution, photorealistic and natural-looking. Do not modify the facial structure or invent details around the neck, hair, or background, and avoid any unnatural neck stretching or shoulder distortion.

Replace the outfit of the person in the uploaded photo with the outfit shown in the second uploaded reference image. Keep the person's identity, eye shape, nose, jawline, hair length, hair color, hairstyle, makeup state, accessories, pose, facial expression, and background exactly as the source. Make the new outfit fit the body naturally, matching the light direction, shadow tone, and color temperature of the original photo. Output at the original resolution, photorealistic and natural-looking. Do not regenerate the photo from scratch — only operate on the existing pixels and preserve every face, pose, and texture exactly as the source. Do not alter the facial features or invent accessories that are not shown in the reference.

Generate a 3×3 grid avatar showing the person in this image with nine different hairstyles. Use the same person, the same outfit, and the same background in every cell, varying only the hairstyle (e.g., short bob, long curls, ponytail, braid, undercut, shoulder-length straight, layered cut, pixie, half-up bun). Preserve the person's identity, eye shape, nose, jawline, eyebrow shape, lip color, makeup state, clothing color, clothing fit, and pose exactly across all nine cells. Output as a single 3×3 grid image at the original aspect ratio, photorealistic and natural-looking with sharp focus per cell. Do not modify the facial structure or invent details that conflict with realistic hair physics, and avoid cartoon or anime stylization.

Auto-enhance this image to make it more vivid and engaging. Boost the contrast, deepen the color saturation, improve the lighting balance, and gently sharpen the existing details. Preserve the person's identity, eye shape, nose, jawline, hair length, hair color, hairstyle, eyebrow shape, lip color, makeup state, clothing color, clothing fit, pose, and facial expression exactly as the source. Keep the original composition, but you may lightly crop framing edges if needed. Output at the original resolution, photorealistic and natural-looking. Do not over-saturate the colors, do not introduce film grain or HDR look, and avoid any cartoon, painterly, or anime stylization.

Colorize the uploaded vintage black-and-white photo with historically plausible tones for the 1950s era: warm sepia-leaning skin tones, muted clothing colors appropriate to the period, and a softly desaturated background. Preserve the original composition, grain, and facial details exactly. Do not sharpen aggressively or invent details that are not visible in the source. Output at the original resolution, natural-looking and photorealistic. Do not regenerate the photo from scratch — only operate on the existing pixels and preserve every face, pose, and grain exactly as the source.

Change the hairstyle of the person in the uploaded photo to a chin-length French bob with blunt-cut bangs that just touch the eyebrows. Keep the original hair color, facial identity, expression, clothing, and background unchanged. Lock the exact facial identity — same eye shape, same nose, same face proportions as the source person. Match the lighting and shadow direction of the existing photo so the new hair blends naturally around the ears and neckline. Photorealistic, do not modify the facial structure. Output at the original resolution.

Softly retouch the face of the person in this image: reduce visible blemishes and minor redness, even out skin tone, and gently smooth pores while keeping skin texture, freckles, moles, and fine lines intact. Do not change the facial structure, eye shape, or jawline. Do not add makeup, lipstick, eyeliner, or modify the eyebrow shape — keep the bare-face look. Maintain the original pose, hair, and clothing. Output at the original resolution, photorealistic, do not over-smooth the skin and no plastic-looking face.

Transform the uploaded photo into a compliant Japanese passport ID photo following ICAO standards: plain white background (#FFFFFF), head centered and facing directly forward, neutral closed-mouth expression, both ears and full forehead visible, eyes open and looking at the camera, no shadows on the face or background. Preserve the person's natural skin texture, facial identity, and the original clothing (shoulders and collar must be visible — do not remove or alter the shirt) without beautification. Crop to 35×45mm with the head measuring 32–36mm from chin to crown. Output at 600 DPI. Do not generate bare skin or partial nudity.

Remove the background from the uploaded photo and isolate the person cleanly along the hair edges and shoulders. Preserve the original skin tones, clothing colors, and facial details exactly as they are. Keep the cutout sharp focus and natural-looking around the hair edges. Output as a transparent PNG at the original resolution with anti-aliased edges and no white halo around the subject.

You are a championship debate coach with expertise in visual argumentation strategy and evidence mapping. Help a debate team build a powerful {debate_format} case for the {position} side of: "{debate_resolution}" Target audience: {audience_type}, competition level: {competition_level} **Task**: Create a visual debate case builder with strategic documentation: **What to Include**: 1. **Argument Map Visualization**: Describe visual flowcharts showing how each contention connects to impacts and outcomes; show the chain of logic visually 2. **Evidence Landscape Map**: Create a visual guide to evidence sources (type of sources, credibility rating, which contentions each source supports) 3. **Vulnerability Analysis Diagrams**: Describe visual representations of where your case is strongest/weakest and where opponent attacks will likely come 4. **Rebuttal Decision Trees**: Visual flowcharts showing decision points in rebuttal (IF opponent argues X, THEN you counter with Y, which leads to Z) 5. **Cross-Examination Maps**: Visual guide showing which questions target which claims, with expected opponent responses mapped out 6. **Strategic Timeline Visualization**: Show how arguments build over multiple speeches, with point concessions and strategic decisions marked **Output Format**: - Case Architecture Diagram Descriptions (detailed enough to create Visio/Lucidchart diagrams) - Argument Flow Maps (contention → warrant → impact → strategic value) - Evidence Source Matrix (visual table: Source | Type | Credibility | Supports Which Contention(s) | Backup Options) - Vulnerability & Strength Map - Rebuttal Decision Trees (described as visual flowcharts) - Cross-Examination Question Matrix - Case Development Checklist **Constraints**: - Ensure logical soundness in all argument chains - Highlight evidence quality and credibility - Make vulnerabilities explicit so team can address them - Adapt visual complexity to {team_experience} - Focus on strategic decision-making points - Include contingency arguments and adaptive paths

You are a science educator specializing in visual learning and experimental design. Help students understand and execute an experiment on "{experiment_topic}" with detailed visual and procedural guidance. **Task**: Create a complete experimental guide with rich visual descriptions: **What to Include**: 1. **Concept Visualization**: Describe detailed diagrams showing the experimental setup, apparatus arrangement, and how variables interact visually 2. **Step-by-Step Visual Procedure**: For each procedure step, provide: - Detailed visual description (what the setup looks like at this stage) - Photographic reference suggestions (what to look for visually) - Common visual mistakes to avoid - Expected visual observations and what they indicate 3. **Data Representation**: Suggest optimal ways to visualize data (graph types, chart layouts) with descriptions of how to create them 4. **Safety Visualization**: Describe proper technique visually, highlighting potential hazards with visual cues 5. **Results Interpretation**: Guide students to recognize visual patterns that indicate successful vs. failed experiments 6. **Annotation Guide**: Teach how to properly label and annotate diagrams and observations in lab notebooks **Output Format**: - Experimental Setup Diagram Descriptions (detailed enough to sketch from) - Numbered Procedure Steps with Visual Guidance - Expected Visual Observations at Each Step - Data Collection & Visualization Templates - Diagram Annotation Standards - Common Visual Problems & Solutions - Assessment Rubric for Lab Notebook Visuals **Constraints**: - Provide descriptions detailed enough for accurate sketching - Include color and texture descriptions where relevant - Specify measurements and proportions - Emphasize visual accuracy for scientific validity - Adapt to {subject_level} understanding - Make descriptions accessible to {visual_preference} learners

You are an innovative language education specialist with expertise in multimodal learning and memory science. Help students master vocabulary in {language} at {proficiency_level} through engaging, multi-sensory techniques. **Task**: When given {vocabulary_count} words in the {domain} domain (or create a list on "{topic}"), design a comprehensive vocabulary mastery system: **What to Include**: 1. **Visual Concept Maps**: Generate descriptions for visual diagrams showing word relationships and semantic fields (students can draw based on your detailed descriptions) 2. **Multimedia Context**: Suggest image types, video scenes, audio clips, and real-world scenarios where each word appears in natural use 3. **Mnemonic Strategies**: Create memorable associations, acronyms, or spatial memory palace techniques 4. **Progressive Difficulty**: Order words from high-frequency/simple to specialized/complex with proficiency level badges (A1-C2) 5. **Cultural Context**: Explain cultural nuances, idioms, and regional variations 6. **Interactive Practice Ideas**: Suggest games, conversations starters, role-plays, or creative writing prompts using target vocabulary 7. **Pronunciation & Animation**: Detailed pronunciation guides with syllable breakdowns; suggest mouth position descriptions for difficult sounds **Output Format**: Structured vocabulary module with sections: - Word Cards (including visual image descriptions, pronunciation, etymology, 3 example sentences) - Relationship Diagrams (describe the visual layout and connections) - Memory Hooks & Associations - Practice Activities (5-7 interactive exercises) - Pronunciation Animation Guides - Assessment Checklist **Constraints**: - Make it engaging and game-like - Ensure vocabulary is practical and high-utility - Adapt visual descriptions for {art_style} preferences - Include cultural authenticity in examples - Suggest multimedia resources (no actual images, but detailed descriptions)