benchmarks.hota.build_mask_tracker
def build_mask_tracker(height: int, width: int, cost_threshold: float = 0.5, class_gate: bool = True) -> unitrack.TrackerBuild a single-stage mask-IoU tracker over a fixed ``(height, width)``.
``cost_threshold`` is the maximum acceptable ``1 - IoU`` association cost,
i.e. matches require ``IoU >= 1 - cost_threshold``. ``NoLifecycle`` keeps
every track (no row drops), which ``ids_per_detection`` relies on.
Score filtering is not done here: low-confidence detections are dropped in
the runner before a ``Detections`` is built, so the tracker only ever sees
kept instances and carries no ``score`` state.
Source: unitrack/benchmarks/hota/tracker.py:49