Overview

Tune2Text is a web application that allows users to extract and transcribe audio from YouTube videos. By simply entering a YouTube video URL, users can obtain a text transcription of the video's audio content.

Setup Instructions

To run the YouTube Transcript App locally, follow these steps:

1. Clone the Repository

2. Install Dependencies Ensure you have Node.js and npm installed. Then, install the required packages:

3. Set Up Environment Variables Create a .env file in the root directory of the project and add your Hugging Face API key: HUGGINGFACE_API_URL="https://api-inference.huggingface.co/models/openai/whisper-small" HUGGINGFACE_API_KEY="your_hugging_face_api_key"

4. Run the Server Start the server using Nodemon for automatic restarts during development: npx nodemon server.js

5. Access the Application Open your web browser and navigate to http://localhost:3000 to access the YouTube Transcript App.

Description of Tools Used

• Hugging Face API: The application utilizes the Hugging Face API, specifically the Whisper model, to perform audio transcription. Whisper is a state-of-the-art speech recognition model that provides high accuracy in transcribing spoken language into text.

• yt-dlp: This tool is used to download audio from YouTube videos. It is a powerful command-line utility that supports a wide range of video and audio formats, making it ideal for extracting audio content for transcription.

• ffmpeg: A multimedia framework used to process audio and video files. In this application, ffmpeg is employed to split the downloaded audio into manageable segments for transcription.