Tone Classification System

This system analyzes audio to detect emotions, including sarcasm and figures of speech.

How to Use

  1. Microphone Input: Record your voice and click 'Analyze Recording'
  2. File Upload: Upload an audio file (.wav or .mp3) and click 'Analyze File'

The system will transcribe the speech, analyze emotions from both text and acoustic features, and display the results with a visualization of emotion probabilities.

About

This tone classification system combines text and acoustic features to detect emotions in speech. It uses a multi-modal approach with:

  • Speech-to-text transcription
  • Text-based emotion analysis
  • Acoustic feature extraction
  • Fusion of both modalities for final prediction

The system can detect: neutral, happy, sad, angry, fearful, disgust, surprised, and sarcastic tones.