tracklab.core package

Submodules

tracklab.core.evaluator module

class tracklab.core.evaluator.Evaluator(cfg)[source]

Bases: ABC

Abstract class to implement for the integration of a new dataset evaluator in wrappers/eval. The functions to implement are __init__ and run. A description of the expected behavior is provided below.

abstract run(tracker_state: TrackerState)[source]: Run the evaluation :param tracker_state: the tracker state for the evaluation :type tracker_state: TrackerState

tracklab.core.visualization_engine module

class tracklab.core.visualization_engine.VisualizationEngine(cfg)[source]

Bases: Callback

after_saved_state = True

draw_frame(image_metadata, detections_pred, ground_truths, image_pred, image_gt, nframes, image=None)[source]

on_image_loop_end(engine: TrackingEngine, image_metadata: Series, image, image_idx: int, detections: DataFrame)[source]

on_video_loop_end(engine, video_metadata, video_idx, detections, image_pred)[source]

on_video_loop_start(engine: TrackingEngine, video_metadata: Series, video_idx: int, index: int)[source]

run(tracker_state: TrackerState, video_id, detections, image_preds, progress=None)[source]