🎬 YT Extractor Pro Ultimate

Intelligent YouTube Slide Extractor, High-Res Media Downloader & AI Study Assistant

A high-performance desktop and web application built with FastAPI, Tailwind CSS, PyWebView, and yt-dlp. Extract high-quality presentation slides from video lectures, download up to 4K video/320kbps MP3s, generate AI video summaries & concept mindmaps, and export study materials to PDF & PPTX.

---

📌 Table of Contents

• Architecture & Workflow
• Key Features
• Tech Stack
• Prerequisites
• Local Installation
• Docker & Cloud Deployment
• Configuration (.env)
• License

---

🏗️ Architecture & Workflow

flowchart LR
    A[YouTube URL] --> B[yt-dlp Engine]
    B --> C[Video Stream Download]
    B --> D[Audio & Subtitles Extract]
    
    C --> E[Scene Transition Detection]
    E --> F[Duplicate Frame Filtering]
    F --> G[Image Sharpener & OCR Filter]
    
    G --> H[Export Deck]
    H --> H1[📄 PDF Document]
    H --> H2[📊 PowerPoint PPTX]
    H --> H3[📝 Markdown Notes]

    D --> I[Offline NLP / Gemini AI Engine]
    I --> J1[🧠 Video Summary]
    I --> J2[🗺️ Concept Mindmap]
    I --> J3[❓ Interactive Q&A]

Loading

---

✨ Key Features

1. 🎞️ Slide Scene Extraction & Enhancement

• Smart Scene Detection: Detects slide changes automatically without capturing duplicate intermediate transition frames.
• Image Enhancer & Text Sharpening: Built-in 1-click filter (ImageEnhance.Contrast + ImageEnhance.Sharpness) that cleans blurry webcam/recording artifacts to make text crystal clear.
• Multi-Format Export: Export cleaned slides directly to:
  • 📄 PDF Presentation (with custom branded title cover page).
  • 📊 PowerPoint Presentation (.pptx).
  • 📦 Zipped JPG/PNG Images.

2. ⚡ Direct Video & Audio Downloader (yt-dlp)

• Resolution Selector: Download up to 4K (2160p), 2K (1440p), 1080p FHD, 720p HD, 480p, and 360p.
• Audio Bitrate Selector: High-fidelity MP3 extraction at 320 kbps, 256 kbps, 192 kbps, and 128 kbps.
• Subtitle Extractor: Download synchronized .srt / .vtt subtitles.

3. 🧠 AI Study Assistant & Dual NLP Engine

• Engine Selector: Switch seamlessly between ⚡ Free Offline NLP (No API Key Required) and 🤖 Google Gemini AI.
• Interactive Concept Mindmaps: Generates visual Mermaid.js mindmaps breaking down core video topics.
• Video Q&A with Timestamp Seeking: Interactive timestamp buttons (e.g., ⏱️ 02:45) that immediately jump the video player to that moment.
• Interactive 3D Study Flashcards: Flip-card memory testing mode with real-time score tracking.

4. 🏷️ Study Tags & Categorization

• Tag slides with color-coded markers (🔴 Exam Alert, 🟢 Definition, 🔵 Code Snippet, 🟡 Key Concept).
• Star favorite slides to create targeted quick-revision decks.

---

🛠️ Tech Stack

• Backend: FastAPI, Uvicorn, Python 3.9+
• Media Engine: yt-dlp, FFmpeg
• Image & Slide Processing: Pillow (PIL), python-pptx, pytesseract OCR
• Frontend / UI: Tailwind CSS, PyWebView, Mermaid.js
• AI / NLP: Google Gemini API / Local Heuristic NLP

---

📦 Prerequisites

1. Python 3.9 or higher installed.
2. FFmpeg:
   • Windows: Auto-downloaded automatically on first run if not found in PATH!
   • Linux / Ubuntu: sudo apt-get install ffmpeg
   • macOS: brew install ffmpeg
3. (Optional) Tesseract OCR for text extraction from slides:
   • Ubuntu: sudo apt-get install tesseract-ocr
   • Windows: Download from UB-Mannheim Tesseract.

---

🚀 Local Installation

1. Clone Repository

git clone https://github.com/Ayush8512/Youtube_Slide_Extractor.git
cd Youtube_Slide_Extractor

2. Set Up Virtual Environment (Recommended)

# Windows
python -m venv venv
venv\Scripts\activate

# Linux / macOS
python3 -m venv venv
source venv/bin/activate

3. Install Dependencies

pip install -r requirements.txt

4. Run the Application

Method A — Run via Python:

python app.py

Method B — Run via Windows Batch Script: Double-click Yt_Slide_Extractor.bat in the project root.

Navigate to http://127.0.0.1:8000 in your browser (or it will automatically launch as a standalone desktop window via PyWebView).

---

🐳 Docker & Cloud Deployment

A production-ready Dockerfile and render.yaml are included for 1-click deployments:

Build & Run with Docker:

docker build -t yt-slide-extractor .
docker run -p 7860:7860 yt-slide-extractor

Open http://localhost:7860.

Works out-of-the-box on Hugging Face Spaces (Docker SDK) and Render.com Web Services.

---

⚙️ Configuration (.env)

Create a .env file in the root directory (refer to .env.example):

# Optional: Needed only if you wish to use the Gemini AI Engine instead of the Free Offline NLP engine
GEMINI_API_KEY=your_gemini_api_key_here

# Server Port (Default: 8000)
PORT=8000

---

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

👤 Author

Ayush Pandey

• GitHub: @Ayush8512
• LinkedIn: Connect on LinkedIn