Skip to content

khyamay/ai-photo-playground

Repository files navigation

AI Photo Restoration Project

Overview

This project is an AI-powered photo restoration application that allows users to upload old or damaged photos and receive restored versions.

Features

  • Upload and restore old or damaged photos
  • AI-powered image enhancement and restoration
  • Download restored images

Tech Stack

  • Frontend: Angular with TypeScript
  • Backend: Express.js with TypeScript
  • Database: PostgreSQL with Prisma ORM
  • AI Model: Replicate API (GFPGAN model)
  • Image Storage: Cloudinary
  • Containerization: Docker

High-Level Architecture

┌─────────────┐     ┌─────────────┐     ┌─────────────┐
│   Angular   │     │  Express.js │     │  PostgreSQL │
│  Frontend   │◄───►│   Backend   │◄───►│  Database   │
└─────────────┘     └─────────────┘     └─────────────┘
      ▲                   ▲                   ▲
      │                   │                   │
      │                   │                   │
      │            ┌──────▼───────┐    ┌──────▼───────┐
      │            │  Replicate   │    │   Prisma     │
      │            │  AI API      │    │     ORM      │
      │            └──────────────┘    └──────────────┘
      │
      │
┌──────▼───────┐
│  Cloudinary  │
│  Storage     │
└──────────────

Upcoming Features

  1. Authentication and Authorization

    • Implement user accounts and login functionality
    • Secure API endpoints with JWT authentication
  2. Image History

    • Allow users to view their previously restored images
  3. Image Management

    • Add the ability for users to delete their old restored images
  4. Enhanced AI Options

    • Provide users with multiple AI model choices for restoration
    • Allow customization of restoration parameters
  5. Batch Processing

    • Enable users to upload and restore multiple images at once

Getting Started

Todo

Contributing

Todo

About

Playground for using AI with Angular

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published