feat(perception): audio scene classifier — detect indoor/outdoor/traffic/park from ambient sound #353

Closed
opened 2026-03-03 13:45:15 -05:00 by sl-jetson · 0 comments
Collaborator

Create a ROS2 node that subscribes to /audio_raw (or microphone via PyAudio). Compute mel-frequency spectral features (MFCC, spectral centroid, bandwidth) over 1-second windows. Classify scene into categories: indoor_quiet, indoor_noisy, outdoor_park, outdoor_traffic, outdoor_wind. Use a simple nearest-centroid classifier with pre-computed reference vectors. Publish /saltybot/audio_scene (String) at 1Hz. Include tests with synthetic audio.

Create a ROS2 node that subscribes to /audio_raw (or microphone via PyAudio). Compute mel-frequency spectral features (MFCC, spectral centroid, bandwidth) over 1-second windows. Classify scene into categories: indoor_quiet, indoor_noisy, outdoor_park, outdoor_traffic, outdoor_wind. Use a simple nearest-centroid classifier with pre-computed reference vectors. Publish /saltybot/audio_scene (String) at 1Hz. Include tests with synthetic audio.
Sign in to join this conversation.
No Label
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: seb/saltylab-firmware#353
No description provided.