SimSearch

Semantic image search: index local photos with CLIP + FAISS, search by text description, browse results in a Flutter UI.

Quick start

1. Backend (Python)

cd app
pip install faiss-cpu torch transformers pillow fastapi uvicorn

Configure folders in app/config.json, then build the index:

python index.py

Start the API server (required for the UI):

python api.py

Or from the repo root: scripts\start_backend.bat

API runs at http://127.0.0.1:8000

• GET /health — index status
• POST /search — body: {"query": "kitchen interior"}

If anything breaks: python clear_db.py then python index.py

2. Frontend (Flutter)

cd frontend\v1
flutter pub get
flutter run -d windows

Ensure the backend is running before searching.

Architecture

images/ + test_media/  →  index.py  →  FAISS + SQLite
                              ↑
User query  →  Flutter UI  →  api.py  →  search.py (CLIP)  →  paths + scores

Benchmarks

Local Processing

CPU

GPU

NPU

NPU Tooling & Compilation:

Cloud Processing (Google Colab - CUDA T4)