Ethanfel 8ab5bdba77 fix: use mean+std MFCC vectors (40-dim) for better discrimination ...

Mean-only vectors were too similar across different audio segments,
causing everything to match even at threshold 0.99. Adding std
captures temporal dynamics and makes the similarity scores much
more spread out.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-04-17 09:27:11 +02:00

..

__init__.py

feat: project skeleton

2026-04-06 11:11:25 +02:00

test_audio_scan.py

fix: use mean+std MFCC vectors (40-dim) for better discrimination

2026-04-17 09:27:11 +02:00

test_utils.py

feat: add apply_keyframes_to_jobs helper

2026-04-14 15:57:27 +02:00