-
-
Detecting negative emotion
-
Shows capacity to recognize multiple people with different expressions
-
Detecting positive emotion
Inspiration
The inspiration behind this project was to use AI to enable people on the autism spectrum to detect emotions better.
What it does
It uses AI to detect the emotions of people on video conferencing calls. It displays their emotions above their heads.
How we built it
We used TensorFlow, pandas, NumPy, and Keras to train the AI. We used OpenCV, MSS, and Flask to build the web application and UI.
Challenges we ran into
We were unable to download TensorFlow and import it into our projects. We had difficulty finding the correct balance between batch size and epoch number to correctly identify emotions.
Accomplishments that we're proud of
The AI is relatively accurate in detecting emotions. This feels like an accomplishment since we have no background in AI.
What we learned
We learned how to use Flask and the OpenCV library. We also learned a lot about training an AI.
What's next for Feelings Finder
It could be made more accurate and the UI could be made more user-friendly.
Log in or sign up for Devpost to join the conversation.