01 Source Code Only
₹99
One-time Payment
- Complete project source code
- Database / data resources
- Dependencies & configuration
- README / setup guide
- Instant download access
NODE-JS Project
Sign Language Recognition System Final Year Project Source Code
Get runnable Sign Language Recognition System Final Year Project Source Code with database files, project setup instructions, live demo options and installation support. This project resource helps students understand the implementation, modules, workflow and technical architecture of a complete NODE-JS project.
Choose the Package That Fits You Best
Simple pricing. Instant access. Secure checkout.
02 Code + Setup Support
₹248
One-time Payment
- Everything in Source Code plan
- Remote setup assistance
- Database & configuration setup
- Run verification
- Help via WhatsApp / Email
Project's Overview
SignBridge — Sign Language to Text & Text to Sign Language System is a full-stack MERN web application developed for sign-language learning, communication, and gesture-based recognition. The application provides two major conversion workflows: Text to Sign and Sign to Text.
In the Text to Sign module, users enter a word or phrase and the application searches its internally managed sign dictionary to display the corresponding sign image or SVG. In the Sign to Text module, users perform a hand sign in front of a webcam. MediaPipe Hand Landmarker extracts hand landmarks in the browser, and a custom k-nearest neighbors (kNN) classifier compares the landmark vector with recognition samples stored in MongoDB to predict the matching sign.
The platform also includes a sign dictionary, categories, favorites, learning mode, webcam-based practice quizzes, conversion history, user profile management, feedback, public sign exploration, admin recognition-sample capture, and a complete administrative control panel.
SignBridge does not depend on a paid external sign-language translation API. The sign dictionary and recognition workflow are maintained internally through database records, uploaded media, and captured hand-landmark training samples.
This project is suitable for B.Tech, M.Tech, BCA, MCA, BE, ME, BSc, MSc, Artificial Intelligence, Machine Learning, Web Development, and Computer Science students who need a practical final year project, major project, or minor project based on sign language recognition, gesture recognition, accessibility, MediaPipe, kNN classification, MERN stack development, webcam processing, and MongoDB.
Login Credentials
Admin
| Field | Value |
|---|---|
[email protected] |
|
| Password | Admin@123 |
Primary User
| Field | Value |
|---|---|
[email protected] |
|
| Password | User@123 |
Related & Popular Projects
View all
Admin Dashboard
- View total users
- View total signs
- View total categories
- View conversions count
- View feedback count
- View recent users
- View recent signs
- View recent feedback
- KPI-style dashboard widgets
Admin User Management
- List users
- Search by name
- Search by email
- Pagination
- View user detail
- Edit name
- Edit mobile
- Activate account
- Deactivate account
- Delete user
- View user's conversion history
Only regular users are shown in the primary user-management list.
Admin Category Management
- View all categories
- View active categories
- View inactive categories
- Create category
- Add category name
- Add description
- Generate/use unique slug
- Edit category
- Activate/deactivate category
- View signs under category
- Delete empty category
Category deletion is blocked when signs still reference that category.
Admin Sign Dictionary Management
- View all signs
- Search signs
- Filter by category
- Pagination
- Create new sign
- Add word
- Add display name
- Select category
- Add description
- Add instructions
- Enable/disable recognition
- Upload sign image
- Upload SVG
- Edit sign
- Activate/deactivate sign
- Replace sign media
- Delete sign
- View signs without media report
Recognition Training Data Management
The administrator controls the training dataset used for Sign-to-Text recognition.
- View recognition-enabled signs
- View recognition sample count per sign
- Find signs with no training samples
- Capture samples using admin webcam
- Extract MediaPipe landmarks
- Send landmark vector to backend
- Save training sample
- List training samples per sign
- Soft-disable sample
- Hard-delete sample
- Improve recognition quality by adding real webcam samples
Recognition samples contain a 63-number hand landmark vector linked to a sign.
Practice Question Management
- List practice questions
- Create practice question
- Select sign
- Select optional category
- Edit practice question
- Delete practice question
- Activate/deactivate question
- Maintain random practice pool
Conversion Management
Admin can monitor conversion activity across the platform.
- List all conversions
- Search conversion
- Filter by Sign-to-Text
- Filter by Text-to-Sign
- Filter by user
- Filter by date range
- Pagination
- Delete conversion
- Monitor practice conversions
Feedback Management
- View all feedback
- Search feedback
- Filter by status
- Pagination
- Update status
- Add admin response
- Mark pending
- Mark reviewed
- Mark resolved
- Delete feedback
Brand & Application Settings
- Update application name
- Update description
- Update contact email
- Update contact number
- Update About text
- Update footer text
- Upload logo
- Store branding image locally
- Update public-site branding
Admin Profile
- Update admin name
- Update admin email
- Change admin password
The complete Admin Panel includes Dashboard, Users, Categories, Sign Dictionary, Recognition Data, Practice Pool, Conversions, Feedback, Brand Settings and Admin Profile modules.
User Registration & Authentication
- User registration
- Name
- Mobile number
- Password
- Confirm password
- Email validation
- Unique email validation
- Minimum password length validation
- User login
- User logout
- JWT authentication
- HTTP-only authentication cookie
- Inactive-user login restriction
User Dashboard
- View active sign count
- View total categories
- View favorite signs count
- View total conversions
- View Sign-to-Text conversion count
- View Text-to-Sign conversion count
- Recent conversion list
- Related sign cards
- Quick navigation to conversion modules
Sign to Text Features
The Sign to Text module lets users perform signs in front of their webcam and convert recognized gestures into text.
Features include:
- Start webcam
- Stop webcam
- Camera permission handling
- Browser-based MediaPipe Hand Landmarker
- Hand landmark extraction
- 63-dimensional landmark vectors
- Load recognition dataset
- kNN classification
- Frame-by-frame predictions
- Stable-label detection over multiple frames
- Display recognized word
- Automatically create conversion history record
- Clear recognition result
- Open conversion history
A stable recognition is stored as a sign_to_text conversion.
Text to Sign Features
The Text to Sign module converts supported text into sign-language media.
- Enter word
- Search sign dictionary
- Normalize search word
- Exact dictionary lookup
- View corresponding sign image/SVG
- View sign details
- Show “not available” when no dictionary entry exists
- Quick-word chips
- Hello
- Bye
- Sorry
- Thank You
- Create conversion history automatically
- Clear lookup state
Successful Text-to-Sign lookups create a text_to_sign conversion record.
Sign Dictionary Features
- Browse active signs
- Search sign by keyword
- Pagination
- View individual sign details
- View sign display name
- View category
- View description
- View instructions
- View sign media
- Add sign to favorites
- Remove sign from favorites
Sign Categories
- Browse active categories
- Open category detail
- View signs under selected category
- Paginated category-specific sign listing
Learn Module
- Browse sign-language content
- Learn signs from dictionary
- Open individual sign details
- Follow guided learning flow
- Practice using the same sign content managed by admin
Practice Module
The Practice module provides webcam-based sign exercises.
- Load random active practice question
- Display expected sign
- Start webcam
- Perform sign
- Run MediaPipe recognition
- Compare predicted word with expected answer
- Mark answer correct/incorrect
- Store practice attempt
- Create conversion record with practice metadata
- Load next question
- Show current session score
The practice pool is controlled by the administrator.
Conversion History
- View personal conversion history
- Search conversion records
- Filter by conversion type
- Sign-to-Text filter
- Text-to-Sign filter
- Pagination
- Delete individual conversion
- Clear entire conversion history
- View conversion timestamp
- View recognized/searched word
Favorites
- Add sign to favorites
- View all favorite signs
- Open favorite sign
- Remove favorite
- Manage favorites from dictionary detail or favorites page
Profile Management
- View profile
- Update name
- Update mobile number
- Change password
- Current password verification
- New password validation
Feedback Module
- Submit feedback
- Select feedback type
- Add subject
- Add message
- View previous feedback
- View feedback status
Supported feedback statuses:
- Pending
- Reviewed
- Resolved
Users can also view admin responses when provided.
Sign Recognition Workflow
User Opens Sign to Text
↓
Browser Webcam Starts
↓
MediaPipe Hand Landmarker
↓
Hand Landmarks Extracted
↓
63-D Landmark Vector
↓
Load Recognition Samples
↓
k-Nearest Neighbors Classification
↓
Frame Predictions
↓
Stable Label Across Frames
↓
Recognized Sign Converted to Text
↓
Conversion Saved to History
MediaPipe performs hand-landmark extraction in the browser while the custom kNN classifier compares the resulting landmark vector with recognition samples stored in MongoDB.
Text to Sign Workflow
User Enters Word
↓
Normalize Input
↓
Search Active Sign Dictionary
↓
Matching Sign Found?
↙ ↘
Yes No
↓ ↓
Display Sign Show Not Available
Image / SVG
↓
Save Text-to-Sign
Conversion History
Recognition Technology
The SignBridge recognition system uses:
- MediaPipe Tasks Vision
- MediaPipe Hand Landmarker
- Hand-landmark extraction
- Custom kNN classifier
- 63-dimensional feature vectors
- MongoDB recognition samples
- Browser webcam
- Stability checks across frames
This is a lightweight machine-learning approach designed around controlled sign classes rather than a paid translation service.
Authentication & Security
Authentication
- JWT-based authentication
- HTTP-only
signbridge_tokencookie - User and Admin roles
- Protected React routes
- Protected backend APIs
Password Security
- Passwords hashed using bcrypt
- Cost factor 10
- Current password required for password change
Rate Limiting
Authentication endpoints use rate limiting:
- Maximum 50 login/register requests
- 15-minute IP window
Role-Based Access
- Express authentication middleware
requireRole('admin')- User/admin route separation
- Non-admin rejected from admin interface
- Inactive accounts blocked from authentication
Production Cookie Security
- Secure cookie flag under production environment
- CORS restricted to configured frontend origin
The current project does not implement email verification, forgot-password email recovery, 2FA, or full account lockout beyond rate limiting.
Technology Used
| Layer | Technology |
|---|---|
| Frontend | React 19 |
| Build Tool | Vite 8 |
| Routing | React Router 7 |
| Styling | Tailwind CSS 4 |
| HTTP Client | Axios |
| Recognition | MediaPipe Tasks Vision |
| Hand Detection | MediaPipe Hand Landmarker |
| Classifier | Custom k-Nearest Neighbors |
| Feature Vector | 63-dimensional hand landmarks |
| Backend | Node.js |
| Backend Framework | Express 4 |
| Database | MongoDB |
| ODM | Mongoose 8 |
| Authentication | JWT |
| Password Hashing | bcryptjs |
| Auth Cookie | HTTP-only Cookie |
| Security | Helmet, CORS, express-rate-limit, express-validator |
| Uploads | Multer |
| Storage | Local disk |
| Development | Node Watch + Vite |
The project uses React, Vite, Node.js, Express, MongoDB, Mongoose, MediaPipe, custom kNN classification, JWT, bcrypt, Multer and Tailwind CSS.
MOST IMPORTANT — Download the Data Folder
First, download the required data folder from Google Drive:
https://drive.google.com/file/d/1-1wc_1GRaw6OUlFynQjhykxHljmY4nPS/view?usp=sharing
After downloading:
- Extract the downloaded file if it is compressed.
- Copy the complete
datafolder. - Paste it inside the project's
serverfolder.
The structure should look like:
SignBridge – Sign Language to Text & Text to Sign Language System
│
├── client
│
└── server
├── data
├── node_modules
├── .env
└── ...
Important: The folder must be available as:
server\data
Do not rename or remove files from the data folder.
Step 1 — Open the Project
Open Command Prompt, PowerShell, or the VS Code terminal and run:
cd "C:\Project\MERN Projects\SignBridge – Sign Language to Text & Text to Sign Language System"
Step 2 — Install Backend Dependencies
cd server
npm install
Wait until all backend dependencies are installed successfully.
Step 3 — Install Frontend Dependencies
cd ../client
npm install
Wait until the installation is complete.
Step 4 — Configure Backend Environment
Go back to the server folder:
cd ../server
Create the .env file from .env.example.
Using Command Prompt
copy .env.example .env
Using PowerShell
Copy-Item .env.example .env
Using Git Bash
cp .env.example .env
Now open:
server\.env
and configure the required environment variables.
For local MongoDB, use:
MONGODB_URI=mongodb://127.0.0.1:27017/signbridge
Keep any other required values according to the .env.example file.
Step 5 — Start MongoDB
Make sure MongoDB is installed and running on your computer.
The project uses the following local MongoDB database:
mongodb://127.0.0.1:27017/signbridge
If MongoDB is installed as a Windows service, ensure the service is running before continuing.
Step 6 — Seed Demo Data
From the server folder, run:
npm run seed
After the seed process completes, verify the seeded data:
npm run seed:verify
Make sure both commands complete successfully.
Step 7 — Start the Backend
From the server folder, run:
npm run dev
Backend will run at:
http://localhost:5000
Keep this terminal running.
Step 8 — Start the Frontend
Open a new terminal window.
Go to the project client folder:
cd "C:\Project\MERN Projects\SignBridge – Sign Language to Text & Text to Sign Language System\client"
Then run:
npm run dev
Frontend will run at:
http://localhost:5173
Final URLs
Frontend
http://localhost:5173
Backend
http://localhost:5000
MongoDB
mongodb://127.0.0.1:27017/signbridge
README Included
Setup Support Available
Admin
| Field | Value |
|---|---|
[email protected] |
|
| Password | Admin@123 |
Primary User
| Field | Value |
|---|---|
[email protected] |
|
| Password | User@123 |