TEAM 9 - Huzaifa, Megan, Prishni, Tanreet

Feasibility

The IBM Tone Analyzer is free to use up to a maximum of 4 million characters per month. Our team projects that this is an appropriate capacity for the launch of our product, and would allow for up to 37 people to use the tool regularly. In the future, as we grow and acquire more users, we will look to incorporate advertising on our website to cover the $4.00 additional charge to raise the cap on the Tone Analyzer so that our application can continue to be free to use.

Tech Implementation

The web app is built using html, css and javascript. We use pyaudio to stream the user voice input. The Google Cloud Text-to-Speech API is used to convert the conversation to text. The IBM Tone Analyzer is used to analyze and determine the emotions portrayed in the conversation. Currently, we are able to detect 7 emotions: joy, sadness, fear, anger, analytical, tentative and confident. Based on combinations of these, we have hardcoded the responses that should be provided to the user, and also are used to change the background color which corresponds to each emotion.

Social Impact

Our app aims to make speech more accessible to members of the Autism Spectrum Disorder (ASD) community who have difficulty recognizing and interpreting subtle and indirect aspects of communication associated with emotional prosody. This tool will allow them to communicate with greater comprehension and confidence. Additionally, this tool will help to raise awareness about the impact that ASD can have on a person’s communication skills and how people outside of the ASD community can better accommodate.

Built With

Share this project:

Updates