ted, polished version for your project page:
Inspiration
Human speech carries emotion that text alone can’t capture. I wanted to create an AI that can listen and understand how someone feels not just what they say. That’s how the Human Speech Sentiment Analyzer idea began: an app that detects the sentiment of spoken audio using advanced AI.
What it does
The Human Speech Sentiment Analyzer takes an audio file as input, analyzes the tone and content of the voice, and predicts whether the sentiment is positive, negative, or neutral. It helps interpret the emotional tone behind speech, giving insights into how people feel based on their voice.
How we built it
I built the app using Python, Streamlit, and the AssemblyAI Speech-to-Text and Sentiment Analysis API. The workflow:
The user uploads an audio file.
AssemblyAI transcribes the speech and analyzes sentiment from the audio.
The app displays the predicted sentiment along with visual cues green for positive, red for negative, and transparent for neutral.
Challenges we ran into
Handling audio files that were too short or unclear for accurate sentiment prediction.
Managing API response delays and integrating them smoothly with Streamlit.
Preventing crashes when no sentiment was detected.
Accomplishments that we're proud of
Successfully built an audio-based sentiment analyzer using AssemblyAI.
Integrated the model into a clean, interactive Streamlit app.
Solved common API and runtime issues for smooth performance.
What we learned
How speech sentiment analysis differs from text sentiment analysis.
The power of AssemblyAI’s models for extracting emotional meaning from audio.
How to deploy and visualize AI predictions in real time.
The importance of handling edge cases like low-quality or mixed-emotion speech.
What's next for Human Speech Sentiment Analyzer
Add real-time microphone recording for instant analysis.
Visualize sentiment trends over time in long audio clips.
Integrate with chatbots or virtual assistants to make them emotionally aware.
Log in or sign up for Devpost to join the conversation.