Prompt 1
role: Act as a facial recognition analyst and casting director. Input: Analyze the person in the uploaded reference image (DO NOT TOUCH FACE). Task 1: Deep Feature Extraction Scan the user's face for distinct biometric markers. Analyze: Bone Structure: Cheekbone height, jaw width, and facial symmetry. Features: Eye shape/tilt, nose bridge/tip, and lip morphology. Expression & Vibe: The "energy" of the face (e.g., intense, friendly, stoic). Task 2: Triangulate Matches Identify the Top 3 celebrity matches from your dataset. Match A: The closest overall structural/biometric match. Match B: A strong match based on specific standout features (e.g., identical eyes or jaw). Match C: A "mood" or stylistic match (similar vibe or grooming style). Task 3: Calculate Likeness Scores Assign a percentage match (0-100%) to each of the three celebrities based on feature correlation. Task 4: Generate 4-Panel Output Create a single composite image with a 2x2 Grid Layout titled "TOP CELEBRITY MATCHES". Layout Instructions: Top Left (Panel 1): The Original User Image (Input). Label: "ORIGINAL". Top Right (Panel 2): Celebrity Match A. Label: "[Name] - [XX]%". Bottom Left (Panel 3): Celebrity Match B. Label: "[Name] - [XX]%". Bottom Right (Panel 4): Celebrity Match C. Label: "[Name] - [XX]%". Ensure the celebrity images are high-quality headshots that match the angle of the user's photo as closely as possible for an accurate comparison.
- Engine
- nano-banana
- Dimension
- square_hd
- Pose
- any
- Crop
- any
- Gender
- female
- Ethnicity
- other
- Age
- twenties
