description
Secondary (SGIE) person-attribute classifier for Karterschool — DINOv2 backbone, sigmoid multi-label over 5 attributes (not_using_phone, using_phone, unknown, child, adult), run on person crops from the primary detector. This is the model that writes the `<label>_scores` attributes rules read: using_phone_scores feeds PAR02/PAR03 phone-usage rules; adult_scores / child_scores feed CNT08/CNT09 when the primary is person-only (karter-yolo11). Classifier only — it does not produce boxes.
use cases
- phone-usage detection on tracked persons (score-based, no phone bbox needed)
- adult/child split downstream of a person-only primary detector
- any per-person binary attribute gate over a time window (ratio/voting rules)
classes
not_using_phoneusing_phoneunknownchildadult