Seek Marine

Ever seen a marine creature and wondered what it is? Seek Marine will let you know!

Seek Marine is a full-stack AI-powered marine species identifier that lets you take or upload a photo and instantly receive a prediction. Inspired by iNaturalist’s Seek app, this project is designed for educational and environmental awareness, and currently supports 23 animal species. Eventually, I want to turn this into a tool for deep-sea creature identification!

Built with FastAPI (backend) and React + Vite + TypeScript + Tailwind CSS (frontend).

---

Features

• Upload or take a photo on your device
• Predict 23 marine species using a trained TensorFlow model
• Instant inference via FastAPI backend
• Responsive and intuitive UI with React + Tailwind CSS
• Frontend-backend integration via REST

---

Project Structure

seek-marine/
├── backend/          
│   ├── main.py
│   ├── model_inception.h5
│   ├── requirements.txt
│   ├── utils.py
├── frontend/
│   ├── src/
│   ├── index.html
│   ├── package.json
│   ├── vite.config.ts
│   ├── .env.example
├── .gitignore
├── README.md

---

Local Setup

1. Clone repository

git clone https://github.com/beepboopdylan/seek-marine.git
cd seek-marine

2. Run the backend (FastAPI)

cd backend
python3 -m venv .venv
source .venv/bin/activate   # On Windows: .\.venv\Scripts\activate

pip install -r requirements.txt
uvicorn main:app --reload --port 8000

3. Run the frontend (React + Vite)

cd ../frontend
cp .env.example .env
npm install
npm run dev

Then open in your browser: http://localhost:5173

4. Environment Variables Create a .env file inside the frontend/ folder:

VITE_API_BASE=http://127.0.0.1:8000

Change it to your deployed backend URL if hosting.

Model Info

• Path: backend/model_inception.h5
• Framework: TensorFlow
• Input: Image (auto-preprocessed to 224×224)
• Output: Outputs one of 23 marine species