| license | Apache-2.0 |
| name | wedding-immortalist |
| description | Transform thousands of wedding photos and hours of footage into an immersive 3D Gaussian Splatting experience with theatre mode replay, face-clustered guest roster, and AI-curated best photos per person. Expert in 3DGS pipelines, face clustering, aesthetic scoring, and adaptive design matching the couple's wedding theme (disco, rustic, modern, LGBTQ+ celebrations). Activate on "wedding photos", "wedding video", "3D wedding", "Gaussian Splatting wedding", "wedding memory", "wedding immortalize", "face clustering wedding", "best wedding photos". NOT for general photo editing (use native-app-designer), non-wedding 3DGS (use drone-inspection-specialist), or event planning (not a wedding planner). |
| allowed-tools | Read,Write,Edit,Bash,Grep,Glob,WebFetch |
| category | Lifestyle & Personal |
| tags | ["wedding","memorial","preservation","photography","storytelling"] |
| pairs-with | [{"skill":"photo-content-recognition-curation-expert","reason":"Curate wedding photos"},{"skill":"event-detection-temporal-intelligence-expert","reason":"Detect wedding events"}] |
Wedding Immortalist
Transform wedding photos and video into an eternal, immersive 3D experience. Create living memories that let couples and guests relive the magic forever.
DECISION POINTS
1. Multi-Space Merge Strategy
Input: Multiple venue spaces detected (ceremony, reception, cocktail hour)
├── < 500 photos per space
│ └── Train unified scene (risk: quality loss, benefit: seamless navigation)
├── 500-1500 photos per space
│ └── Train per-space with portal transitions (RECOMMENDED)
├── > 1500 photos per space
│ └── Split large spaces into sub-regions to avoid memory overflow
└── Mixed lighting conditions?
├── Yes → Separate indoor/outdoor scenes
└── No → Can merge with lighting adaptation
2. Face Clustering Threshold Selection
Input: Face embedding similarity scores across all detected faces
├── Photo count < 1000
│ └── Conservative threshold (0.4): Prefer precision over recall
├── Photo count 1000-5000
│ └── Balanced threshold (0.3): Standard wedding size
├── Photo count > 5000
│ └── Aggressive threshold (0.25): Large weddings need looser matching
└── Professional + phone photos mixed?
├── Yes → Increase threshold by 0.05 (lighting/quality variance)
└── No → Use base threshold
3. Theme Signal Conflict Resolution
Input: Competing aesthetic signals detected
├── Venue theme ≠ Decor theme
│ ├── Venue rustic, decor modern → Weight decor 70%, venue 30%
│ └── Venue formal, decor casual → Weight venue 60%, decor 40%
├── Multiple cultural traditions
│ └── Create fusion palette prioritizing couple's primary heritage
├── Seasonal vs. chosen colors
│ ├── Spring wedding with fall colors → Honor chosen colors
│ └── Christmas wedding ignoring season → Blend seasonal + chosen
└── Guest attire contradicts theme
└── Ignore guest attire, focus on venue + intentional decor
4. Video Frame Extraction Rate
Input: Wedding video files with metadata
├── Handheld/phone video (high motion blur)
│ └── Extract 1 fps to reduce blur frames
├── Professional video (smooth)
│ └── Extract 2-3 fps for optimal overlap
├── Drone footage
│ └── Extract 1 fps (smooth motion, redundant frames)
├── Ceremony (mostly static)
│ └── Extract 1 fps, boost for processional/recessional
└── Reception/dancing (high motion)
└── Extract 3 fps to capture movement variety
5. Space Skip Threshold Decision
Input: Reconstructed space quality metrics
├── < 50 images AND reconstruction error > 5.0 pixels
│ └── Skip space: insufficient coverage, warn user
├── 50-100 images AND error > 3.0 pixels
│ └── Include with quality warning, mark as "preview quality"
├── > 100 images AND error > 2.0 pixels
│ └── Retrain with adjusted parameters, likely lighting issues
└── Blurry photos > 40% of space images
└── Skip space, request better photos from user
