Adora ML Core Model is a lightweight yet powerful AI engine designed for automated creative validation, layout correction, and final ad rendering. Built for hackathons and real-world marketing automation, it integrates OCR, object detection, banned-phrase compliance, safe-zone checking, and auto-layout optimization into one unified pipeline โก. The system intelligently analyzes product packshots, detects design issues, and applies auto-fixes to produce compliant, visually appealing creativesโready for campaigns or downstream workflows. With a clean FastAPI backend and a modern Streamlit UI ๐จ, this project offers an efficient end-to-end solution for automated creative intelligence, ensuring accuracy, speed, and usability for both developers and non-technical users ๐.
- ๐ Smart Creative Validation โ OCR, YOLO detections, banned-phrase scanning, and brand-compliance logic
- ๐จ AI Auto-Fix Engine โ Adjusts font size, color, contrast, and safe-zone placement automatically
- ๐ง Background Removal โ Clean, transparent product cutouts using robust segmentation
- โก FastAPI Backend โ High-performance API endpoints for validate, autofix, render & health check
- ๐ผ Render Engine โ Composes final creatives with text, product packshots, and backgrounds
- ๐ Modern Streamlit Frontend โ User-friendly UI for testing, previewing, and interacting with the system
- ๐ Modular Architecture โ Easy to extend, plug-and-play components for hackathons or production use
python -m venv .venv .\.venv\Scripts\Activate.ps1 pip install -r requirements.txt uvicorn app.main:app --reload --host 0.0.0.0 --port 8000 streamlit run streamlit_app.py winget install -e --id UB-Mannheim.TesseractOCR Get-ChildItem "C:\Program Files" -Recurse -Filter tesseract.exe -ErrorAction SilentlyContinue tesseract --version