Pierre: The Presentation Assistant

Pierre is a smart service that analyzes your speech patterns/ body language and helps you become a stronger communicator.

Comment

Inspiration

Many speakers have bad habits that can take away from their message while presenting. We wanted to create a tool that could help fix that.

What it does

Pierre helps people improve their presentation skills. It uses machine learning to analyze a presenters dialog flow live and gives them notifications when they fall victim to bad habits such as poor posture or filler words.

How we built it

We built the service on several libraries. We used OpenCV (for video), PocketSphinx (for live NLP), and Standard Library to connect everything.

Challenges we ran into

We ran into many challenges regarding the GUI of our project. We were unable to create a graphical interface by the deadline so we pivoted and decided to use Tkinter as the main provider of information.

Accomplishments that we're proud of

We are proud of our speech detection and SMS messages. We are also very proud of our posture detection.

What we learned

We used and tried to use many tools that we have never touched before. One of the main ones was Standard Library's tools. We had no idea that it existed and that it was so powerful.

What's next for Pierre: The Presentation Assistant

We will be conducting more user research to help us with our product direction. We will also continue to work on it in our spare time. The next big update will be the addition of a GUI that allows users to view their live data as they are presenting and also compare their recent sessions with older ones to see if they have improved.

Built With

Share this project:

Updates