Subash Kumar Yadav
HomeAboutProjectsBlogContact

Subash Kumar Yadav

Full-stack developer building web and mobile products that are useful, reliable, and thoughtfully designed.

Explore

  • Projects
  • Blog
  • About
  • Contact

Connect

  • GitHub
  • DAKSHA
  • subash67335@gmail.com
  • LinkedIn
© 2026–present Subash Kumar Yadav. Built with Next.js and thoughtful product design.
Back to projects

Case study

Colorizer

Web AppIn development

This project focuses on converting black-and-white images into vibrant color using an AI-driven approach. The system utilizes a convolutional neural network (CNN) based autoencoder and a conditional Generative Adversarial Network (cGAN) to generate realistic colorized images.

Repository

Problem

The challenge of colorizing black-and-white images is a complex task that requires understanding the context and semantics of the image to produce accurate and visually appealing colors. Traditional methods often fall short in capturing the nuances of different objects and scenes, leading to unrealistic or dull colorizations.

My implementation

Developed the AI model architecture, trained the networks on a diverse dataset of black-and-white and color image pairs, and implemented the web interface for users to upload images and receive colorized outputs.

Core features

  • AI-driven image colorization using CNN and cGAN
  • User-friendly web interface for image upload and colorization
  • Real-time processing and display of colorized images
  • Support for various image formats and resolutions
  • Backend API for handling image processing requests
  • Integration with a database for storing user-uploaded images and results

Stack

Next.jsTypeScriptTailwind CSSMongoDBNode.jsExpress.jsReactPyTorchDjango

Use cases & project notes

The minor-project is an application designed to colorize grayscale images using advanced artificial intelligence. It provides a user-friendly interface where individuals can upload their black-and-white photos and retrieve vibrantly colored versions. Key features include secure user authentication, personal image history management to track past conversions, and a powerful deep learning model that performs the actual colorization.

Visual Overview

View diagram source
flowchart TD
    A0["User Authentication System
"]
    A1["Image Colorization Service
"]
    A2["Frontend Application Structure
"]
    A3["Image History Management
"]
    A4["Deep Learning Colorization Model
"]
    A2 -- "Manages access via" --> A0
    A0 -- "Secures operations for" --> A1
    A0 -- "Secures data for" --> A3
    A2 -- "Initiates tasks with" --> A1
    A1 -- "Utilizes" --> A4
    A1 -- "Populates data for" --> A3
    A2 -- "Presents and interacts with" --> A3

Under the Hood: How a Page is Displayed

Let's trace what happens from the moment you type your application's address (like your-app.com/) into your browser:

View diagram source
sequenceDiagram
    participant User
    participant App.tsx
    participant Protected.tsx
    participant Layout.tsx
    participant HomePage.tsx

    User->>App.tsx: Navigates to / (Home page)
    App.tsx->>Protected.tsx: Checks if user is logged in for / route
    Note over Protected.tsx: If not logged in, redirects to login page.
    Protected.tsx->>App.tsx: User is logged in (continues rendering)
    App.tsx->>Layout.tsx: Renders Layout with HomePage as children
    Layout.tsx->>TopBar.tsx: Renders the TopBar
    Layout.tsx->>Sidebar.tsx: Renders the Sidebar
    Layout.tsx->>HomePage.tsx: Places HomePage content in main area
    HomePage.tsx->>User: Displays interactive content
    User-->>App.tsx: Sees fully structured Home page

Project evidence

Verified from the local project folder atd GitHub repository metadata for subash-0/minor-project.

Current status

In development

Project links

Repository