"LaZyxEWU team on AMD AI League Hackathon"
Team Idea
SnapIdentity automates event photography: raw photos auto-upload, get AI-enhanced (color correction, cropping, quality fixes), and are instantly matched to guests via selfie face-scan retrieval — no manual editing, no scrolling through shared galleries. Built as the AI core of a larger multi-tenant photographer SaaS platform. Face embeddings and auto-edit models run on AMD GPUs via ROCm-accelerated PyTorch for fast batch processing of large event photo sets. Stack: FastAPI backend, InsightFace/OpenCV/GFPGAN for AI, FAISS/pgvector for similarity search, S3 storage, React/Next.js frontend. Impact: near-zero photographer editing time and an instant, private photo gallery for every guest.
.png&w=256&q=75)
Submission
No submission yet