Draw a polygon on a scene. Args: scene: The scene to draw the polygon on. polygon: The polygon to be drawn, given as a list of ve

Draws a rounded rectangle on an image. Args: scene: The image on which the rounded rectangle will be drawn. rect: The rectan

Calculate confusion matrix for a batch of detections for a single image. Args: predictions: Batch prediction. Describes

Fill the sequence with padding elements until the sequence reaches the desired size. Args: sequence: The input sequence.

Filters a list of polygons based on their area. Args: polygons: A list of polygons, where each polygon is represented by

Create a Color instance from a BGR tuple. Args: color_tuple: A tuple representing the color in BGR format, where each

Parse and scale bounding boxes and masks from Google Gemini 2.5 style [JSON output](https://ai.google.dev/gemini-api/docs/vision?lang=python)

Parse and rescale bounding boxes and class labels from Qwen-2.5-VL JSON output. The JSON is expected to be enclosed in triple backticks with

Creates a Detections instance from a [YOLO-NAS](https://github.com/Deci-AI/super-gradients/blob/master/YOLONAS.md) inference

Generator function to yield frames from an RTSP stream. Args: rtsp_url (str): URL of the RTSP video stream. Yields: np.

Converts a hex color string (e.g. "#FF00FF" or "#FF00FF80") to an RGBA tuple. Args: hex_color: A hex color string. Returns:

Joins two lists of tracks, ensuring that the resulting list does not contain tracks with duplicate internal_track_id values. Args:

List files in a directory with specified extensions or all files if no extensions are provided. Args: directory: The directo

Merges two Detections objects into a single Detections object. Assumes each Detections contains exactly one object. A `winning` detectio

Get the current ID and increment the counter. Returns: The newly assigned ID.

Overlay image onto scene at specified anchor point. Handles cases where overlay position is partially or completely outside scene bounds.

Resize image to specified resolution. Can optionally maintain aspect ratio. Args: image: The image to resize. resolution_wh:

* Strips Python REPL prompts (>>> and ...) from code text. * Also removes output lines (lines that don't start with >>> or ...). * * NOTE: This is

Adds a new time stamp to the deque for FPS calculation.

Splits the data into two parts using the provided train_ratio. Args: data: The data to split. train_ratio: The ratio of the

Run Kalman filter correction step. Args: mean: The predicted state's mean vector. covariance: The state's co

Converts bounding box coordinates from `(x, y, width, height)` format to `(x_min, y_min, x_max, y_max)` format. Args: xywh: A nu

Args: color: The color or color palette to use for annotating detections. thickness: Thickness of the

Compute the axis-aligned bounding envelope of each oriented box. Args: polygons: ``(N, 4, 2)`` array of polygon corners. Returns:

Adjusts the position of labels to ensure they stay within the frame boundaries. Args: resolution_wh: The width and heigh

Compute derived summary columns from scenario results. Returns a DataFrame with all ScenarioResult fields plus derived columns (ratios, speed

Build a CompactMask using tight per-mask bounding boxes.

Broadcastable function, computing the F1 score from the confusion matrix. Args: confusion_matrix: shape (N, ..., 3), whe

Broadcastable function, computing the precision from the confusion matrix. Args: confusion_matrix: shape (N, ..., 3), wh

Broadcastable function, computing the recall from the confusion matrix. Args: confusion_matrix: shape (N, ..., 3), where

Return boxes, masks or oriented bounding boxes from detections.

Return boxes, masks or oriented bounding boxes from detections.

Return boxes, masks or oriented bounding boxes from detections.

Return boxes, masks or oriented bounding boxes from detections.

Parse annotator tab structure from docs/detection/annotators.md. Returns: Mapping of category name to list of annotator tab labels within

Return class names from a YOLO data.yaml file ordered by class index. Supports list and dict forms of the ``names`` field. Dict keys that are

Return a copy of detections with contents filtered by object size.

Return a copy of detections with contents filtered by object size.

Return a copy of detections with contents filtered by object size.

Return a copy of detections with contents filtered by object size.

Generate bounding boxes defining the coordinates of image slices with overlap. Args: resolution_wh: Image resolution `(w

Generate random bounding boxes within given image dimensions and size constraints. Creates `count` bounding boxes randomly positioned and si

Greedy non-maximum merging loop, independent of how overlap is computed. ``iou_against_candidate(order_indices, candidate_idx)`` must return the

Apply greedy version of non-maximum merging to avoid detecting too many overlapping bounding boxes for a given object. Args: pre

Apply greedy version of non-maximum merging to avoid detecting too many Args: predictions: An array of shape `(n, 5)` containing

Build a single-detection Detections with iscrowd=0 from a (1, H, W) mask.

Merge a group of single-object Detections into one merged detection. Used internally by :meth:`Detections.with_nmm` to combine each merge group

Merge multiple OBB corner arrays using winner-angle projection. The first entry in *corners_list* is the winner. Its orientation angle (deriv

Greedy NMS suppression loop given a precomputed pairwise IoU matrix. Assumes `ious` is square with row/column order matching `categories`. De

Dispatch NMM grouping per class, then translate local indices back to the global row positions of ``predictions``. ``group_within(global_indi

Compute the area of each oriented-box polygon using the shoelace formula. Args: polygons: ``(N, 4, 2)`` array of polygon corners. Re

Add an agnostic class column when missing, sort by descending score. Returns the score-descending sort index, the reordered predictions, and the

Nearest-neighbour resize of (N, H, W) bool masks to (N, new_h, new_w). Uses floor-division indexing (``arange * src // dst``) to match the st

Return the i-th element when the value stores per-detection data. Dispatch rules: - np.ndarray with ndim == 0: return as

Validate detection-level confidence: 1D ``np.ndarray`` with shape ``(n,)``.

Validates that the number of provided labels matches the number of detections. Args: labels: A list of labels, one for each detectio

Factory function to create synthetic YOLO-format datasets with custom parameters. Args: tmp_path: Pytest tmp_path fixture nu

Annotate a frame with zone and box annotations based on given detections. Args: frame (np.ndarray): The original frame to be annotat

Annotate a frame with zone and box annotations based on given detections. Args: frame (np.ndarray): The original frame to be annotat

Annotates the given scene with oriented bounding boxes based on the provided detections. Args: scene: The image

Annotates the given scene with box masks based on the provided detections. Args: scene: The image where bounding boxes w

Annotates the given scene with ellipses based on the provided detections. Args: scene: The image where ellipses will be

Annotates the given scene with box corners based on the provided detections. Args: scene: The image where box corners wi

Annotates the given scene with circles based on the provided detections. Args: scene: The image where box corners will b

Annotates the given scene with dots based on the provided detections. Args: scene: The image where dots will be drawn.

Annotates the given scene with labels based on the provided detections, with support for Unicode characters. Args:

Annotates the given scene with triangles based on the provided detections. Args: scene: The image where triangles will b

Annotates the given scene with bounding boxes with rounded edges based on the provided detections. Args: scene:

Annotates the given scene with percentage bars based on the provided detections. The percentage bars visually represent the confidenc

Annotates the provided scene with scaled and cropped parts of the image based on the provided detections. Each detection is cropped f

Highlights the differences between two sets of detections. Args: scene: The image where detections will be drawn.

Returns the index map of source classes -> target classes.

Compute the average precision using 101-point interpolation (COCO), given the recall and precision curves. Args:

Provides a generator that yields chunks of the input sequence of the size specified by the `batch_size` parameter. The last chunk may be