Video/Audio Summarization Application

A Python-based Streamlit Video/Audio Summarization Application that allows users to upload audio or video files, transcribe the content, and generate concise summaries for easy reference.

Features

• Upload and process video/audio files for transcription
• Automatic transcription of audio content using Wav2Vec 2.0
• Summarize lengthy transcriptions into brief, digestible content
• User-friendly interface with real-time processing and feedback

Screenshots

Home Screen
Transcription Output
Summary Output

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/VideoAudioSummarizationApp.git
   cd VideoAudioSummarizationApp

2. Install dependencies:

   pip install -r requirements.txt

3. Run the Streamlit app:

   streamlit run app.py

Usage

1. Open the application in your browser (typically at http://localhost:8501).

2. Upload a video or audio file in .mp4, .wav, or .mp3 format.

3. The app will extract audio (if a video file is uploaded), transcribe it, and display the transcription.

4. View the summarized content in the Summary section.

Adding Audio/Video Samples to GitHub

To add sample audio or video files to GitHub:

1. Place sample files in a directory within the project, such as sample_files/.
2. In your README, provide links to these files for easy access.
3. Use these sample files for demo purposes or to facilitate testing and contributions.

Built With

• Python - The programming language used.
• Streamlit - For the interactive web application.
• Hugging Face Transformers - For speech-to-text and summarization models.
• Librosa - For audio processing.

Contributing

Contributions are welcome! To contribute, please submit a pull request and follow the standard GitHub workflow.

Acknowledgments

• Hugging Face community for providing state-of-the-art NLP models.
• Inspiration from various NLP resources for implementing the summarization feature.

Author

Lav Kalsi