Calculating detections dwell time in zones, using video file. Args: zone_configuration_path: Path to the zone configuration JSON fil

Interactively draw polygons on images or video frames and save the annotations. Args: source_path: Path to the source image or video

Script to stream videos using RTSP protocol. Args: video_directory: Directory containing video files to stream. number_of_st

Download a specific YouTube video by providing its URL. Args: url: The full URL of the YouTube video you wish to download. o

Video Processing with Inference and ByteTrack. Args: source_video_path: Path to the source video file target_video_path: Pat

Video Processing with YOLO and ByteTrack. Args: source_weights_path: Path to the source weights file source_video_path: Path

Vehicle Speed Estimation using Inference and Supervision. Args: source_video_path: Path to the source video file target_vide

Vehicle Speed Estimation using Ultralytics and Supervision. Args: source_video_path: Path to the source video file target_vi

Vehicle Speed Estimation using YOLO-NAS and Supervision. Args: source_video_path: Path to the source video file target_video

Counting people in zones with Inference and Supervision. Args: zone_configuration_path: Path to the zone configuration JSON file

Counting people in zones with YOLO and Supervision. Args: zone_configuration_path: Path to the zone configuration JSON file

the mAP score at IoU threshold of `0.5`.

the mAP score at IoU thresholds from `0.5` to `0.95`.

the mAP score at IoU threshold of `0.75`.

Deprecated. Use `supervision.detection.utils.converters.mask_to_rle`.

Merge a list of `DetectionDataset` objects into a single `DetectionDataset` object. This method takes a list of `Detecti

Concatenate multiple :class:`CompactMask` objects into one. All inputs must have the same ``image_shape``. Args: masks_l

Given N detections each of length 1 (exactly one object inside), combine them into a single detection object of length 1. The contained inner

Given N detections each of length 1 (exactly one object inside), combine them into a single detection object of length 1. The contained inner

Mock callback function for testing.

Return per-mask crop origins as ``(x1, y1)`` integer offsets. Returns: Array of shape ``(N, 2)`` with ``int32`` offsets.

Plot the Mean Average Recall results. ![example_plot](\ https://media.roboflow.com/supervision-docs/metrics/mAR_plot_exa

Create confusion matrix plot and save it at selected location. Args: save_path: Path to save the plot. If not provided,

Plot the mAP results. ![example_plot]( https://media.roboflow.com/supervision-docs/metrics/mAP_plot_example.png

Plot the recall results. ![example_plot]( https://media.roboflow.com/supervision-docs/metrics/recall_plot_example.png

Plot the F1 results. {

Plot the precision results. ![example_plot]( https://media.roboflow.com/supervision-docs/metrics/precision_plot_example.

Plots image using matplotlib. Args: image: The frame to be displayed ImageType is a flexible type, accepting either `nu

Plots images in a grid using matplotlib. Args: images: A list of images as ImageType is a flexible type, accepting eithe

Clears all the time stamps from the deque.

Reset internal metric state.

Reset the metric to its initial state, clearing all stored data.

Reset the metric to its initial state, clearing all stored data.

Resets the internal state of the ByteTrack tracker. This method clears the tracking data, including tracked, lost, and remov

Deprecated. Use `supervision.detection.utils.converters.rle_to_mask`.

Save image to target directory with optional custom filename. Args: image: Image to save with shape `(height, width, 3)`

Return ``(N, H, W)`` matching the dense mask convention. Returns: Tuple ``(N, H, W)``. Examples: ```pycon

Splits the dataset into two parts (training and testing) using the provided split_ratio. Args: split_ratio:

Splits the dataset into two parts (training and testing) using the provided split_ratio. Args: split_ratio:

2x downscale of a 4x4 block mask halves dimensions.

2x upscale of a 2x2 mask doubles each pixel.

Assert every annotator in EXPECTED_ANNOTATOR_TAB_GROUPS has a tab in the docs. Tests flat membership only — does not enforce which category each

All-False mask resizes to all-False regardless of dimensions.

All-False 3x3 splits into three columns each [3].

Zero-area masks should produce IoU = 0, not NaN.

All-True mask resizes to all-True regardless of dimensions.

All-True 2x2 splits into two columns each [0, 2].

Test download_assets when file already exists and is valid.

Test download_assets when file exists but is corrupted (re-downloads).