ALBIREO

Detector-agnostic, training-free adaptive inference for video object detection on edge GPUs

Video object detection on the edge runs an expensive detector on every frame, even when consecutive frames are nearly identical — yet naive frame skipping is content-blind and fails exactly when it matters: object entry, occlusion recovery, abrupt motion.

ALBIREO lets uncertainty decide. It wraps off-the-shelf detectors, tracks each object with a 10-dimensional Kalman filter, and invokes the real detector only when prediction uncertainty exceeds a threshold; skipped frames get predicted boxes at near-zero GPU cost. A rescue mechanism preserves confirmed objects through brief detector misses, and an empty-scene screen avoids detector calls on objectless frames. No detector modification or retraining is required.

Evaluated on BDD100K MOT across three architecturally distinct detectors (YOLO11x, YOLO26x, RF-DETR-Large) and two Jetson platforms (AGX Thor, AGX Orin):

  • AP@50 stays within ±1.2 pp of per-frame inference while total energy drops 12.1–17.6%.
  • On the primary YOLO26x configuration, AP@50 improves by +0.8 pp while energy falls 17.6% (Thor), with per-frame energy-delay product reduced by ~25%.
  • A fixed-interval skipping baseline with the same skip budget loses 8.6 pp AP@50.

The paper will appear at ACM/IEEE SEC 2026 (Taherin et al., 2026).

References

2026

  1. SEC
    Amir Taherin, José Cano, Bin Ren, and 2 more authors
    In ACM/IEEE Symposium on Edge Computing (SEC). To appear , 2026