Track Objects (BoostTrack)

How it fits

Typical backends

Caveats

• I/O contractOutput is track-centric. Each frame emits the set of active CONFIRMED tracks, so the count is the number of live tracks, not the number of input detections.
• I/O contractThe [TrackedSegmentation] output arm is only selectable when the input is [Segmentation]; the carried mask comes from the input. With [BoundingBox] input it cannot be resolved and the worker errors at startup.
• I/O contractThe [Track] arm is the only one that fills each track's ``trajectory``; it holds the track's own past centre points, oldest first.
• AccuracyBoostTrack's edge is crowded, occluded scenes - it boosts each detection's confidence and inflates soft-BIoU overlap before the track_thresh gate, so an occluded object on a weak detection is kept and matched rather than dropped.
• AccuracyAssociation is motion (Kalman) plus boosted geometry only, with no appearance model. Add reidentify_objects_torchreid downstream when ids must survive long occlusions or identical-looking objects crossing.
• State lifetimeA new track is tentative until matched on min_hits consecutive frames, then confirmed and emitted. A confirmed track unmatched for up to track_buffer frames coasts on its Kalman prediction; after that it is dropped and its id is never reused. State is in-process, so a restart resets all ids.
• Parameter interactionmatch_thresh is the IoU-distance gate - a track/detection pair whose 1 - IoU exceeds it cannot link. track_thresh is the high-confidence gate for the first association pass and for spawning new tracks; the confidence and soft-BIoU boosting is applied before it. class_agnostic false confines matches to the same class id.