Mental State Detector

Real-time voice analysis to infer emotional and mental states from pitch, volume, tempo, and timbre.

Voice Input

Waveform will appear here when you start speaking

How it works

1. Click Start and allow microphone access.

2. Speak naturally — the tool analyzes your voice in real time.

3. Pitch is detected via autocorrelation; volume via RMS amplitude.

4. Spectral centroid measures voice brightness; ZCR approximates speech rate.

5. A softmax classifier maps these features to six mental states.

Disclaimer

This is a research demonstration using heuristics on acoustic features. It is not a clinical diagnostic tool. For mental health concerns, please consult a qualified professional.