Skip to content

ronikdedhia/Sign-Language-Converter

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Sign Language Converter

Python Django OpenCV YOLOv5 ResNet-50 License

An end-to-end American Sign Language (ASL) platform that works in both directions: it recognises ASL gestures from video/camera with 92% accuracy using ResNet-50 / YOLOv5, and converts English text into ASL animations using NLP preprocessing and Blender-rendered sign videos. Delivered as a Django web application.


Features

  • ASL Recognition — Gesture detection using YOLOv5 and ResNet-50 (92 % accuracy)
  • Text-to-Sign Generation — Enter any English sentence; it is NLP-processed (tokenized, POS-tagged, stop-word removed, lemmatized) and converted to a sequence of ASL sign videos
  • Tense detection (past, present, future, continuous) to apply grammar-correct sign ordering
  • 200+ pre-rendered Blender sign-video assets (letters A–Z, digits 0–9, common words and phrases)
  • User authentication: sign-up, login, logout (Django Auth)
  • Pages: Home, About, Contact, Index (sign dashboard), Animation player
  • Reduces translation time by 25% compared to manual interpretation

Tech Stack

Component Technology
Language Python 3
Web Framework Django 3.0
CV / Recognition OpenCV, YOLOv5, ResNet-50
NLP NLTK (tokenize, POS tag, stopwords, WordNetLemmatizer)
Sign Assets Blender-rendered .mp4 animations
Auth Django built-in authentication
Frontend HTML/CSS + custom animation player

Getting Started

Prerequisites

pip install django opencv-python nltk torch torchvision
python -m nltk.downloader punkt averaged_perceptron_tagger stopwords wordnet omw-1.4

1. Clone the repository

git clone https://github.com/ronikdedhia/Sign-Language-Converter.git
cd Sign-Language-Converter

2. Apply migrations

python manage.py migrate

3. Run the development server

python manage.py runserver

Open http://127.0.0.1:8000/ in your browser.

Demo — Text to Sign

  1. Log in or create an account
  2. Navigate to http://127.0.0.1:8000/animation/
  3. Type an English sentence and submit
  4. Watch the corresponding ASL sign video sequence play automatically

File Structure

├── A2SL/
│   ├── settings.py          # Django settings
│   ├── urls.py              # URL routing (home, about, contact, login, signup, animation)
│   ├── views.py             # NLP pipeline + animation logic
│   └── wsgi.py / asgi.py
├── assets/
│   ├── *.mp4                # 200+ Blender sign videos (A-Z, 0-9, words)
│   └── CSS/                 # Stylesheets + animations
├── templates/               # HTML templates
│   ├── home.html
│   ├── index.html
│   ├── about.html
│   ├── contact.html
│   └── ...
├── manage.py
└── LICENSE

URL Routes

URL View Description
/ home_view Landing page
/index/ index_view Sign language dashboard
/animation/ animation_view Text → ASL animation (auth required)
/signup/ signup_view User registration
/login/ login_view User login
/logout/ logout_view User logout
/about/ about_view About page
/contact/ contact_view Contact page

Screenshots

App Interface

Logo Landing Page Chat Interface Text to Sign

Logo  ·  Landing Page  ·  Chat Interface  ·  Text to Sign

Text to Sign — Output Sign to Text Sign to Text — Output Hand Tracking

Text to Sign — Output  ·  Sign to Text  ·  Sign to Text — Output  ·  Hand Tracking

Dataset & Research

ASL Dataset ASL Dataset — Detail Dataset Overview Dataset Preprocessing

ASL Dataset  ·  ASL Dataset — Detail  ·  Dataset Overview  ·  Dataset Preprocessing

Image Transformation Lighting Conditions Model Specs

Image Transformation  ·  Lighting Conditions  ·  Model Specs

License

This project is licensed under the MIT License.

About

ASL recognition (92% acc, ResNet+YOLOv5) and text-to-sign generation (NLP + Blender animations) via a Django web app.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors