# Analysis Module

Computer vision and audio analysis utilities.

## Planned Components

### motion.py
Simple motion detection using frame differencing.
- Compare consecutive frames
- Threshold for "significant" motion
- Returns bounding boxes of movement

### audio_classify.py  
Audio event classification using YAMNet.
- Runs on Orin (GPU accelerated)
- Classifies: speech, dog bark, door, music, etc.
- Returns event type + confidence

### vad.py
Voice Activity Detection.
- Silero VAD or energy-based
- Filters silence before sending to classifier
- Reduces bandwidth and processing