We've always found it frustrating to seek through long videos such as those for presidential debates. This is why we created Vibez, which intelligently analyzes videos and provides summaries in the form of emotional profiles and keywords.
Specifically, we built a web application that takes any YouTube video and graphs the major sentiments in the speech over time. The application also finds important keywords in the video and allows the users to search for locations mentioning specific words or phrases.
We built our application using Flask as the backend, Bootstrap as the frontend, and IBM Watson for speech-to-text and emotion analysis.
Run python app/app.py
in terminal to start the web server