🤟 SignBridge: Making Gestures Accessible

Bridging Communication Gaps with Real-Time Gesture Translation

---

✨ At a Glance

Category	Details
Purpose	Real-time sign language translation for video calls
Tech Stack	Python, TensorFlow/MediaPipe, OBS, JavaScript,
Deployment	Standalone EXE + OBS Virtual Camera
Key Features	Gesture recognition, dynamic overlays, TTS
Team	4 developers (see below)

---

🌟 Why SignBridge?

"Because every gesture deserves to be seen and heard."

We solve three critical challenges in virtual communication:

1-👁️ Visibility Issues - Clear text overlays that stay visible for all participants
2-🔄 Platform Fragmentation - Works with any video conferencing tool via OBS Virtual Cam
3-⏱️ Real-Time Lag - Instant gesture-to-text translation with <200ms latency

---

Technical Deep Dive

---

🌐 Website Preview

---

👥 Meet The Team

Name	Role	Contributions	LinkedIn
Arjit	Core Developer	• Webcam/OBS pipeline • TTS integration • Cross-platform testing	LinkedIn
Apoorv	ML Engineer	• EXE packaging • Model training • Performance optimization	LinkedIn
Ishita	UI/UX Designer	• Website development • Branding • EXE download portal	LinkedIn
Sree	QA & Docs	• Testing • Documentation • OBS scene templates	LinkedIn

---

🖼️ Preview Exe

---

🚀 Getting Started

💿 Installation (For Users)

1. Download 👉 SignBridgePro_Setup.exe
2. Install OBS Studio + Virtual Camera plugin
3. Run SignBridge and follow the setup wizard

---

🔧 Manual Build (For Developers)

This section is for developers who want to build, modify, or contribute to SignBridgePro locally. It includes complete details of the build pipeline, folder structure, and troubleshooting tips.

---

📁 Project Structure

1.Model Working logic/
├── model/                              # Trained model files
│   ├── label_map.npy
│   └── sign_model.h5
├── test_sequence/                      # Folder for test input sequences
├── generate_test_sequence.py           # Creates test image sequences
├── predict_from_images.py              # Predicts from static image folders
├── requirements.txt                    # Model-specific dependencies
├── SETUP.txt                           # Setup or instructions (optional)
├── simulate_sequence_video.py          # Simulates sign input from video
├── test_model.py                       # Model evaluation or testing script
└── train_model.ipynb                   # Notebook to train the sign recognition model

2.exe logic formation/
├── assets/                             # App icons and visuals
│   └── signBridge_icon.ico
├── build/                              # Auto-generated temp files during PyInstaller build
│   └── SignBridgePro/
├── dependencies/                       # System-level dependencies
│   └── VC_redist.x64.exe               # Required for running EXE on Windows
├── dist/                               # Final build artifacts
│   └── caption_output.txt              # Output from real-time translations
├── installers/                         # OBS installers for virtual camera support
│   ├── OBS-Studio-31.1.1-Windows-x64-installer.exe
│   └── OBS-VirtualCam2.0.4-installer.exe
├── model/                              # ML model and label mappings
│   ├── label_map.npy
│   └── sign_model.h5
├── build_exe.bat                       # Script to generate EXE from Python
├── create_installer.bat                # Script to generate Windows installer
├── LICENSE.TXT
├── main.py                             # Main Python script for real-time translation
├── README.md
├── requirements.txt                    # Python dependencies
├── settings.json                       # App config (e.g., voice toggle, overlay path)
├── SignBridgePro.spec                  # Auto-generated PyInstaller config
└── simple_installer.nsi               # NSIS script for installer creation

3.Website Code/
├── .gitignore                        # Git ignore rules
├── apoorv.jpg                        # Project asset
├── arijit.jpg                        # Project asset
├── GitHub_logo.png                   # GitHub icon
├── hand.jpg                          # Hand icon
├── Hocurx_ppt.pdf                    # Presentation
├── image.png                         # Web graphic
├── index.html                        # Main HTML entry point
├── ishita.jpg                        # Project asset
├── linkedin_logo.png                 # LinkedIn icon
├── logo_sb.jpg                       # Project logo
├── package.json                      # Node.js dependencies
├── package-lock.json                 # Exact dependency tree
├── script.js                         # Frontend JavaScript
├── server.js                         # Backend (Node.js) script
├── sree.jpg                          # Project asset
├── styles.css                        # Custom styling
└── .bolt/                            # Tool or config folder (purpose-defined)

---

🛠️ Build Process Explained

1. 🔁 Clone & Setup the Project

git clone https://github.com/your-repo/SignBridge.git
cd SignBridge
pip install -r requirements.txt

2. ⚙️ Generate Executable

build_exe.bat

This does the following:

1. Generates SignBridgePro.spec

2. Creates build/SignBridgePro/ folder (temp build files)

3. Creates dist/ folder with final .exe and dependencies

3. 📦 Create Windows Installer

create_installer.bat

This uses simple_installer.nsi with NSIS to:

1. Bundle the .exe from dist/

2. Include dependencies (OBS, VirtualCam, VC_redist)

3. Generate SignBridgePro_Setup.exe

---

💡 Pro Tips for Smooth Setup

✅ First-Time Setup

1. Install NSIS from: https://nsis.sourceforge.io/
2. Install OBS Studio + Virtual Camera Plugin
3. If EXE gives DLL errors → run VC_redist.x64.exe from dependencies/

---

🐛 Debugging Builds

1. Check build/warn-SignBridgePro.txt if the build fails
2. Ensure all necessary files are present in the dist/ folder
3. OBS might lock files — close OBS before running build_exe.bat

---

🛠 Customizing the Installer

Edit the simple_installer.nsi script to:

1. Change the default installation directory
2. Add uninstall shortcut or registry entries
3. Embed a license agreement, changelog, or version notes