feat: Voice message distress analysis — paralinguistic features #131

Open
opened 2026-04-15 15:15:06 +00:00 by Rockachopa · 2 comments
Owner

Epic: #102 (Multimodal Crisis Detection)

Task

Analyze voice messages (OGG on Telegram) for distress signals using paralinguistic features.

Signals to Detect

  • Speech rate changes (very slow or very fast)
  • Pitch variability reduction (monotone = depression indicator)
  • Long pauses / silence ratio
  • Vocal tremor / shakiness
  • Volume drops

Implementation

# Use whisper for transcription + feature extraction
# Or: dedicated paralinguistic model
def analyze_voice_message(audio_path: str) -> dict:
    return {
        "transcript": str,
        "speech_rate": float,       # words per minute
        "pitch_mean": float,
        "pitch_variability": float,
        "silence_ratio": float,     # 0-1
        "distress_score": float     # 0-1
    }

Deliverables

  • Voice analysis module: the-door/voice_analysis.py
  • Integration with Telegram voice message handler
  • Distress score thresholds: low (<0.3), medium (0.3-0.7), high (>0.7)
  • Test: 10 voice samples, verify score correlates with content

Labels: multimodal, voice, crisis, priority:high

## Epic: #102 (Multimodal Crisis Detection) ### Task Analyze voice messages (OGG on Telegram) for distress signals using paralinguistic features. ### Signals to Detect - Speech rate changes (very slow or very fast) - Pitch variability reduction (monotone = depression indicator) - Long pauses / silence ratio - Vocal tremor / shakiness - Volume drops ### Implementation ```python # Use whisper for transcription + feature extraction # Or: dedicated paralinguistic model def analyze_voice_message(audio_path: str) -> dict: return { "transcript": str, "speech_rate": float, # words per minute "pitch_mean": float, "pitch_variability": float, "silence_ratio": float, # 0-1 "distress_score": float # 0-1 } ``` ### Deliverables - [ ] Voice analysis module: `the-door/voice_analysis.py` - [ ] Integration with Telegram voice message handler - [ ] Distress score thresholds: low (<0.3), medium (0.3-0.7), high (>0.7) - [ ] Test: 10 voice samples, verify score correlates with content ### Labels: multimodal, voice, crisis, priority:high
hermes was assigned by Rockachopa 2026-04-16 01:21:39 +00:00
hermes was unassigned by Rockachopa 2026-04-17 05:03:42 +00:00
hermes was assigned by Rockachopa 2026-04-17 05:10:59 +00:00
Author
Owner

🔥 BURN DISPATCH — mac:WARD.10

Hundred-Pane Fleet active. Working on it.

🔥 **BURN DISPATCH** — mac:WARD.10 Hundred-Pane Fleet active. Working on it.
Author
Owner

🔥 BURN DISPATCH — mac:WARD.10

Hundred-Pane Fleet active. Working on it.

🔥 **BURN DISPATCH** — mac:WARD.10 Hundred-Pane Fleet active. Working on it.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Timmy_Foundation/the-door#131