An inability to communicate well often creates barriers that are nearly impossible to take down.

SpeakEasy was built to help people improve their communicaiton skills by recording a video of them speaking, and then using deep learning and computer vision to give them useful feedback on where they can improve.

• Fenil Doshi, Rohit Saha - worked on the deep learning and computer vision scripts using Jupyter Notebook
• Samuel Zhang, Tyler Jiang - built the web interface with HTML, CSS, JavaScript, Flask, and Python

Checkout a limited demo of the project here: Front-end Demo

The user records themselves speaking (as if they are presenting or pitching something), then presses the "Analyze" button.

The application performs an analysis of the video, telling you what your emotions were like, as well as places where you could switch up your word choice.

We had some issues integrating the video analysis scripts with the web app itself, so the application currently displays the analysis of one of our pre-recorded videos.

The landing page (inspired by the Stripe design).