NLP / Audio ML✓ CompletedOctober 2024

AI Voice Turing Test Model

Developed for the Social Bias Hackathon hosted by Ethical Spectacle Research & ASU. Built a voice-based classification system that identifies human vs. robotic calls using audio feature extraction and NLP-based sentiment understanding. The project included an interactive evaluation demo and bias analysis of classification behavior across demographic groups — earning 3rd place recognition.

Tech Stack

PythonTransformersNLPAudio ProcessingLibrosaHuggingFace

Key Highlights

  • 🏆 3rd Place — ASU Social Bias Hackathon
  • Audio feature extraction: MFCCs, pitch, cadence, spectral features
  • NLP sentiment analysis for transcript-based signals
  • Bias analysis across demographic categories
  • Interactive demo for real-time call classification