Automatic Video Summarization

AGA - Aitolkyn Baigutanova, Gabriel Lima, Assem Zhunis

Automatic Video Summarization is an interactive platform for automatic summarization of lectures.

This project was generated with React 17.0.1.

Prototype link: https://bit.ly/3waOHcd

Instructions

1. Run the back end:

   • cd /flash-backend

   • In a new Anaconda environment:

     • pip install flask-cors
     • pip install waitress
     • pip install youtube_transcript_api
     • pip install wikipedia
     • pip install bert-extractive-summarizer
     • install PyTorch - depends on OS and CUDA/CPU - https://pytorch.org/

   • python app.py

2. Run the front end:
   • cd /react-frontend
   • npm start
3. Choose a video with subtitles
   • Examples: https://www.youtube.com/watch?v=cUS_22_lDiM https://www.youtube.com/watch?v=R9OCA6UFE-0

Classes/Pages/Scripts

• flash-backend/app.py - Flask server that receives all requests, and returns summaries and Wikipedia information.
• react-frontend/src/App.js - Controls which page is shown.
• react-frontend/src/Main.js - Main page with video on the left and summary on the right.
• react-frontend/src/Edit.js - Page where full transcription is shown and users can navigate through the text.
• react-frontend/src/Input.js - Initial page.

Dependencies

1. Python
   • Anaconda, flask-cors, waitress, youtube_transcript_api, wikipedia, bert-extractive-summarizer, pytorch
2. React
   • React, Bootstrap, jQuery, react-loader-spinner, react-youtube, react-promise-tracker, react-native-yt-player, react-player, react-contenteditable