🌈 Prism — AI-Powered Computer Vision & Nutrition Inference Engine

Prism is a production-grade Computer Vision and Nutrition Analytics API. It combines state-of-the-art instance segmentation, depth-based volumetric estimation, and neural dish classification to deliver precise macro and micro-nutrient breakdowns from simple meal photographs.

---

🏗️ Architecture & ML Pipeline

Prism executes a multi-stage inference pipeline on every incoming image request:

[ User Image ]
      │
      ├──>  1. YOLO Segmentation (foodseg_best.pt) ──> Mask & Area Extraction
      ├──>  2. Depth & Geometry Engine         ──> Volume & Mass Estimation (g)
      ├──>  3. Food-101 Classifier             ──> Plate-level Categorization
      │
      └──>  4. USDA Mapping Engine             ──> Complete Nutrients (Macros & Micros)
                  │
                  └──> [ JSON Payload Response ]

Key Technical HighlightsLazy-Loaded Models: Models are loaded lazily to eliminate weight reloading overhead across FastAPI requests.Volumetric Mass Estimation: Calculates relative depth and bounding geometry to estimate food mass in grams.Granular Micronutrients: Provides detailed vitamin, mineral, lipid, and caloric metrics scaled to estimated mass.🚀 Quick Start1. PrerequisitesPython 3.10+Virtual Environment (.venv)2. InstallationBash# Clone the repository git clone https://github.com/Divansu47/Prism.git cd Prism

Set up virtual environment

python -m venv .venv source .venv/bin/activate # On Windows: .venv\Scripts\activate

Install dependencies

pip install -r requirements.txt

3. Local Development ServerBashuvicorn main:app --reload The API server will run locally at http://127.0.0.1:8000.

---

📡 API Reference

POST /analyze-food

Analyze a food photograph to extract volumetric metrics and full nutritional values.

Request (Multipart Form-Data)

Parameter	Type	Description
file	File (Binary)	JPEG or PNG image of the meal

Example cURL Request

curl -X POST "https://food-nutrition-920479426497.asia-south1.run.app/analyze-food" \
  -F "file=@/path/to/meal_image.jpeg"

---

☁️ DeploymentPrism is optimized for containerized deployments on Google Cloud Run.Bashgcloud run deploy food-nutrition
--source .
--region asia-south1
--allow-unauthenticated
--memory 4Gi
--cpu 2
--cpu-boost 📜 LicenseDistributed under the Apache License 2.0. See LICENSE for more details.

How to update it in VS Code:

1. Open C:\Games\food-nutition-ml\README.md.
2. Delete everything inside it and paste the raw code block above.
3. Save the file (Ctrl + S).
4. In your terminal, run:

   git add README.md
   git commit -m "docs: fix markdown formatting and code blocks in README"
   git push origin main

   ---