Skip to content

Reference for ultralytics/utils/metrics.py

Note

This file is available at https://github.com/ultralytics/ultralytics/blob/main/ultralytics/utils/metrics.py. If you spot a problem please help fix it by contributing a Pull Request ūüõ†ÔłŹ. Thank you ūüôŹ!


ultralytics.utils.metrics.ConfusionMatrix

ConfusionMatrix(nc, conf=0.25, iou_thres=0.45, task='detect')

A class for calculating and updating a confusion matrix for object detection and classification tasks.

Attributes:

Name Type Description
task str

The type of task, either 'detect' or 'classify'.

matrix ndarray

The confusion matrix, with dimensions depending on the task.

nc int

The number of classes.

conf float

The confidence threshold for detections.

iou_thres float

The Intersection over Union threshold.

Source code in ultralytics/utils/metrics.py
def __init__(self, nc, conf=0.25, iou_thres=0.45, task="detect"):
    """Initialize attributes for the YOLO model."""
    self.task = task
    self.matrix = np.zeros((nc + 1, nc + 1)) if self.task == "detect" else np.zeros((nc, nc))
    self.nc = nc  # number of classes
    self.conf = 0.25 if conf in {None, 0.001} else conf  # apply 0.25 if default val conf is passed
    self.iou_thres = iou_thres

matrix

matrix()

Returns the confusion matrix.

Source code in ultralytics/utils/metrics.py
def matrix(self):
    """Returns the confusion matrix."""
    return self.matrix

plot

plot(normalize=True, save_dir='', names=(), on_plot=None)

Plot the confusion matrix using seaborn and save it to a file.

Parameters:

Name Type Description Default
normalize bool

Whether to normalize the confusion matrix.

True
save_dir str

Directory where the plot will be saved.

''
names tuple

Names of classes, used as labels on the plot.

()
on_plot func

An optional callback to pass plots path and data when they are rendered.

None
Source code in ultralytics/utils/metrics.py
@TryExcept("WARNING ‚ö†ÔłŹ ConfusionMatrix plot failure")
@plt_settings()
def plot(self, normalize=True, save_dir="", names=(), on_plot=None):
    """
    Plot the confusion matrix using seaborn and save it to a file.

    Args:
        normalize (bool): Whether to normalize the confusion matrix.
        save_dir (str): Directory where the plot will be saved.
        names (tuple): Names of classes, used as labels on the plot.
        on_plot (func): An optional callback to pass plots path and data when they are rendered.
    """
    import seaborn  # scope for faster 'import ultralytics'

    array = self.matrix / ((self.matrix.sum(0).reshape(1, -1) + 1e-9) if normalize else 1)  # normalize columns
    array[array < 0.005] = np.nan  # don't annotate (would appear as 0.00)

    fig, ax = plt.subplots(1, 1, figsize=(12, 9), tight_layout=True)
    nc, nn = self.nc, len(names)  # number of classes, names
    seaborn.set_theme(font_scale=1.0 if nc < 50 else 0.8)  # for label size
    labels = (0 < nn < 99) and (nn == nc)  # apply names to ticklabels
    ticklabels = (list(names) + ["background"]) if labels else "auto"
    with warnings.catch_warnings():
        warnings.simplefilter("ignore")  # suppress empty matrix RuntimeWarning: All-NaN slice encountered
        seaborn.heatmap(
            array,
            ax=ax,
            annot=nc < 30,
            annot_kws={"size": 8},
            cmap="Blues",
            fmt=".2f" if normalize else ".0f",
            square=True,
            vmin=0.0,
            xticklabels=ticklabels,
            yticklabels=ticklabels,
        ).set_facecolor((1, 1, 1))
    title = "Confusion Matrix" + " Normalized" * normalize
    ax.set_xlabel("True")
    ax.set_ylabel("Predicted")
    ax.set_title(title)
    plot_fname = Path(save_dir) / f'{title.lower().replace(" ", "_")}.png'
    fig.savefig(plot_fname, dpi=250)
    plt.close(fig)
    if on_plot:
        on_plot(plot_fname)

print

print()

Print the confusion matrix to the console.

Source code in ultralytics/utils/metrics.py
def print(self):
    """Print the confusion matrix to the console."""
    for i in range(self.nc + 1):
        LOGGER.info(" ".join(map(str, self.matrix[i])))

process_batch

process_batch(detections, gt_bboxes, gt_cls)

Update confusion matrix for object detection task.

Parameters:

Name Type Description Default
detections Array[N, 6] | Array[N, 7]

Detected bounding boxes and their associated information. Each row should contain (x1, y1, x2, y2, conf, class) or with an additional element angle when it's obb.

required
gt_bboxes Array[M, 4] | Array[N, 5]

Ground truth bounding boxes with xyxy/xyxyr format.

required
gt_cls Array[M]

The class labels.

required
Source code in ultralytics/utils/metrics.py
def process_batch(self, detections, gt_bboxes, gt_cls):
    """
    Update confusion matrix for object detection task.

    Args:
        detections (Array[N, 6] | Array[N, 7]): Detected bounding boxes and their associated information.
                                  Each row should contain (x1, y1, x2, y2, conf, class)
                                  or with an additional element `angle` when it's obb.
        gt_bboxes (Array[M, 4]| Array[N, 5]): Ground truth bounding boxes with xyxy/xyxyr format.
        gt_cls (Array[M]): The class labels.
    """
    if gt_cls.shape[0] == 0:  # Check if labels is empty
        if detections is not None:
            detections = detections[detections[:, 4] > self.conf]
            detection_classes = detections[:, 5].int()
            for dc in detection_classes:
                self.matrix[dc, self.nc] += 1  # false positives
        return
    if detections is None:
        gt_classes = gt_cls.int()
        for gc in gt_classes:
            self.matrix[self.nc, gc] += 1  # background FN
        return

    detections = detections[detections[:, 4] > self.conf]
    gt_classes = gt_cls.int()
    detection_classes = detections[:, 5].int()
    is_obb = detections.shape[1] == 7 and gt_bboxes.shape[1] == 5  # with additional `angle` dimension
    iou = (
        batch_probiou(gt_bboxes, torch.cat([detections[:, :4], detections[:, -1:]], dim=-1))
        if is_obb
        else box_iou(gt_bboxes, detections[:, :4])
    )

    x = torch.where(iou > self.iou_thres)
    if x[0].shape[0]:
        matches = torch.cat((torch.stack(x, 1), iou[x[0], x[1]][:, None]), 1).cpu().numpy()
        if x[0].shape[0] > 1:
            matches = matches[matches[:, 2].argsort()[::-1]]
            matches = matches[np.unique(matches[:, 1], return_index=True)[1]]
            matches = matches[matches[:, 2].argsort()[::-1]]
            matches = matches[np.unique(matches[:, 0], return_index=True)[1]]
    else:
        matches = np.zeros((0, 3))

    n = matches.shape[0] > 0
    m0, m1, _ = matches.transpose().astype(int)
    for i, gc in enumerate(gt_classes):
        j = m0 == i
        if n and sum(j) == 1:
            self.matrix[detection_classes[m1[j]], gc] += 1  # correct
        else:
            self.matrix[self.nc, gc] += 1  # true background

    if n:
        for i, dc in enumerate(detection_classes):
            if not any(m1 == i):
                self.matrix[dc, self.nc] += 1  # predicted background

process_cls_preds

process_cls_preds(preds, targets)

Update confusion matrix for classification task.

Parameters:

Name Type Description Default
preds Array[N, min(nc, 5)]

Predicted class labels.

required
targets Array[N, 1]

Ground truth class labels.

required
Source code in ultralytics/utils/metrics.py
def process_cls_preds(self, preds, targets):
    """
    Update confusion matrix for classification task.

    Args:
        preds (Array[N, min(nc,5)]): Predicted class labels.
        targets (Array[N, 1]): Ground truth class labels.
    """
    preds, targets = torch.cat(preds)[:, 0], torch.cat(targets)
    for p, t in zip(preds.cpu().numpy(), targets.cpu().numpy()):
        self.matrix[p][t] += 1

tp_fp

tp_fp()

Returns true positives and false positives.

Source code in ultralytics/utils/metrics.py
def tp_fp(self):
    """Returns true positives and false positives."""
    tp = self.matrix.diagonal()  # true positives
    fp = self.matrix.sum(1) - tp  # false positives
    # fn = self.matrix.sum(0) - tp  # false negatives (missed detections)
    return (tp[:-1], fp[:-1]) if self.task == "detect" else (tp, fp)  # remove background class if task=detect





ultralytics.utils.metrics.Metric

Metric()

Bases: SimpleClass

Class for computing evaluation metrics for YOLOv8 model.

Attributes:

Name Type Description
p list

Precision for each class. Shape: (nc,).

r list

Recall for each class. Shape: (nc,).

f1 list

F1 score for each class. Shape: (nc,).

all_ap list

AP scores for all classes and all IoU thresholds. Shape: (nc, 10).

ap_class_index list

Index of class for each AP score. Shape: (nc,).

nc int

Number of classes.

Methods:

Name Description
ap50

AP at IoU threshold of 0.5 for all classes. Returns: List of AP scores. Shape: (nc,) or [].

ap

AP at IoU thresholds from 0.5 to 0.95 for all classes. Returns: List of AP scores. Shape: (nc,) or [].

mp

Mean precision of all classes. Returns: Float.

mr

Mean recall of all classes. Returns: Float.

map50

Mean AP at IoU threshold of 0.5 for all classes. Returns: Float.

map75

Mean AP at IoU threshold of 0.75 for all classes. Returns: Float.

map

Mean AP at IoU thresholds from 0.5 to 0.95 for all classes. Returns: Float.

mean_results

Mean of results, returns mp, mr, map50, map.

class_result

Class-aware result, returns p[i], r[i], ap50[i], ap[i].

maps

mAP of each class. Returns: Array of mAP scores, shape: (nc,).

fitness

Model fitness as a weighted combination of metrics. Returns: Float.

update

Update metric attributes with new evaluation results.

Source code in ultralytics/utils/metrics.py
def __init__(self) -> None:
    """Initializes a Metric instance for computing evaluation metrics for the YOLOv8 model."""
    self.p = []  # (nc, )
    self.r = []  # (nc, )
    self.f1 = []  # (nc, )
    self.all_ap = []  # (nc, 10)
    self.ap_class_index = []  # (nc, )
    self.nc = 0

ap property

ap

Returns the Average Precision (AP) at an IoU threshold of 0.5-0.95 for all classes.

Returns:

Type Description
(ndarray, list)

Array of shape (nc,) with AP50-95 values per class, or an empty list if not available.

ap50 property

ap50

Returns the Average Precision (AP) at an IoU threshold of 0.5 for all classes.

Returns:

Type Description
(ndarray, list)

Array of shape (nc,) with AP50 values per class, or an empty list if not available.

curves property

curves

Returns a list of curves for accessing specific metrics curves.

curves_results property

curves_results

Returns a list of curves for accessing specific metrics curves.

map property

map

Returns the mean Average Precision (mAP) over IoU thresholds of 0.5 - 0.95 in steps of 0.05.

Returns:

Type Description
float

The mAP over IoU thresholds of 0.5 - 0.95 in steps of 0.05.

map50 property

map50

Returns the mean Average Precision (mAP) at an IoU threshold of 0.5.

Returns:

Type Description
float

The mAP at an IoU threshold of 0.5.

map75 property

map75

Returns the mean Average Precision (mAP) at an IoU threshold of 0.75.

Returns:

Type Description
float

The mAP at an IoU threshold of 0.75.

maps property

maps

MAP of each class.

mp property

mp

Returns the Mean Precision of all classes.

Returns:

Type Description
float

The mean precision of all classes.

mr property

mr

Returns the Mean Recall of all classes.

Returns:

Type Description
float

The mean recall of all classes.

class_result

class_result(i)

Class-aware result, return p[i], r[i], ap50[i], ap[i].

Source code in ultralytics/utils/metrics.py
def class_result(self, i):
    """Class-aware result, return p[i], r[i], ap50[i], ap[i]."""
    return self.p[i], self.r[i], self.ap50[i], self.ap[i]

fitness

fitness()

Model fitness as a weighted combination of metrics.

Source code in ultralytics/utils/metrics.py
def fitness(self):
    """Model fitness as a weighted combination of metrics."""
    w = [0.0, 0.0, 0.1, 0.9]  # weights for [P, R, mAP@0.5, mAP@0.5:0.95]
    return (np.array(self.mean_results()) * w).sum()

mean_results

mean_results()

Mean of results, return mp, mr, map50, map.

Source code in ultralytics/utils/metrics.py
def mean_results(self):
    """Mean of results, return mp, mr, map50, map."""
    return [self.mp, self.mr, self.map50, self.map]

update

update(results)

Updates the evaluation metrics of the model with a new set of results.

Parameters:

Name Type Description Default
results tuple

A tuple containing the following evaluation metrics: - p (list): Precision for each class. Shape: (nc,). - r (list): Recall for each class. Shape: (nc,). - f1 (list): F1 score for each class. Shape: (nc,). - all_ap (list): AP scores for all classes and all IoU thresholds. Shape: (nc, 10). - ap_class_index (list): Index of class for each AP score. Shape: (nc,).

required
Side Effects

Updates the class attributes self.p, self.r, self.f1, self.all_ap, and self.ap_class_index based on the values provided in the results tuple.

Source code in ultralytics/utils/metrics.py
def update(self, results):
    """
    Updates the evaluation metrics of the model with a new set of results.

    Args:
        results (tuple): A tuple containing the following evaluation metrics:
            - p (list): Precision for each class. Shape: (nc,).
            - r (list): Recall for each class. Shape: (nc,).
            - f1 (list): F1 score for each class. Shape: (nc,).
            - all_ap (list): AP scores for all classes and all IoU thresholds. Shape: (nc, 10).
            - ap_class_index (list): Index of class for each AP score. Shape: (nc,).

    Side Effects:
        Updates the class attributes `self.p`, `self.r`, `self.f1`, `self.all_ap`, and `self.ap_class_index` based
        on the values provided in the `results` tuple.
    """
    (
        self.p,
        self.r,
        self.f1,
        self.all_ap,
        self.ap_class_index,
        self.p_curve,
        self.r_curve,
        self.f1_curve,
        self.px,
        self.prec_values,
    ) = results





ultralytics.utils.metrics.DetMetrics

DetMetrics(save_dir=Path('.'), plot=False, on_plot=None, names=())

Bases: SimpleClass

This class is a utility class for computing detection metrics such as precision, recall, and mean average precision (mAP) of an object detection model.

Parameters:

Name Type Description Default
save_dir Path

A path to the directory where the output plots will be saved. Defaults to current directory.

Path('.')
plot bool

A flag that indicates whether to plot precision-recall curves for each class. Defaults to False.

False
on_plot func

An optional callback to pass plots path and data when they are rendered. Defaults to None.

None
names tuple of str

A tuple of strings that represents the names of the classes. Defaults to an empty tuple.

()

Attributes:

Name Type Description
save_dir Path

A path to the directory where the output plots will be saved.

plot bool

A flag that indicates whether to plot the precision-recall curves for each class.

on_plot func

An optional callback to pass plots path and data when they are rendered.

names tuple of str

A tuple of strings that represents the names of the classes.

box Metric

An instance of the Metric class for storing the results of the detection metrics.

speed dict

A dictionary for storing the execution time of different parts of the detection process.

Methods:

Name Description
process

Updates the metric results with the latest batch of predictions.

keys

Returns a list of keys for accessing the computed detection metrics.

mean_results

Returns a list of mean values for the computed detection metrics.

class_result

Returns a list of values for the computed detection metrics for a specific class.

maps

Returns a dictionary of mean average precision (mAP) values for different IoU thresholds.

fitness

Computes the fitness score based on the computed detection metrics.

ap_class_index

Returns a list of class indices sorted by their average precision (AP) values.

results_dict

Returns a dictionary that maps detection metric keys to their computed values.

curves

TODO

curves_results

TODO

Source code in ultralytics/utils/metrics.py
def __init__(self, save_dir=Path("."), plot=False, on_plot=None, names=()) -> None:
    """Initialize a DetMetrics instance with a save directory, plot flag, callback function, and class names."""
    self.save_dir = save_dir
    self.plot = plot
    self.on_plot = on_plot
    self.names = names
    self.box = Metric()
    self.speed = {"preprocess": 0.0, "inference": 0.0, "loss": 0.0, "postprocess": 0.0}
    self.task = "detect"

ap_class_index property

ap_class_index

Returns the average precision index per class.

curves property

curves

Returns a list of curves for accessing specific metrics curves.

curves_results property

curves_results

Returns dictionary of computed performance metrics and statistics.

fitness property

fitness

Returns the fitness of box object.

keys property

keys

Returns a list of keys for accessing specific metrics.

maps property

maps

Returns mean Average Precision (mAP) scores per class.

results_dict property

results_dict

Returns dictionary of computed performance metrics and statistics.

class_result

class_result(i)

Return the result of evaluating the performance of an object detection model on a specific class.

Source code in ultralytics/utils/metrics.py
def class_result(self, i):
    """Return the result of evaluating the performance of an object detection model on a specific class."""
    return self.box.class_result(i)

mean_results

mean_results()

Calculate mean of detected objects & return precision, recall, mAP50, and mAP50-95.

Source code in ultralytics/utils/metrics.py
def mean_results(self):
    """Calculate mean of detected objects & return precision, recall, mAP50, and mAP50-95."""
    return self.box.mean_results()

process

process(tp, conf, pred_cls, target_cls)

Process predicted results for object detection and update metrics.

Source code in ultralytics/utils/metrics.py
def process(self, tp, conf, pred_cls, target_cls):
    """Process predicted results for object detection and update metrics."""
    results = ap_per_class(
        tp,
        conf,
        pred_cls,
        target_cls,
        plot=self.plot,
        save_dir=self.save_dir,
        names=self.names,
        on_plot=self.on_plot,
    )[2:]
    self.box.nc = len(self.names)
    self.box.update(results)





ultralytics.utils.metrics.SegmentMetrics

SegmentMetrics(save_dir=Path('.'), plot=False, on_plot=None, names=())

Bases: SimpleClass

Calculates and aggregates detection and segmentation metrics over a given set of classes.

Parameters:

Name Type Description Default
save_dir Path

Path to the directory where the output plots should be saved. Default is the current directory.

Path('.')
plot bool

Whether to save the detection and segmentation plots. Default is False.

False
on_plot func

An optional callback to pass plots path and data when they are rendered. Defaults to None.

None
names list

List of class names. Default is an empty list.

()

Attributes:

Name Type Description
save_dir Path

Path to the directory where the output plots should be saved.

plot bool

Whether to save the detection and segmentation plots.

on_plot func

An optional callback to pass plots path and data when they are rendered.

names list

List of class names.

box Metric

An instance of the Metric class to calculate box detection metrics.

seg Metric

An instance of the Metric class to calculate mask segmentation metrics.

speed dict

Dictionary to store the time taken in different phases of inference.

Methods:

Name Description
process

Processes metrics over the given set of predictions.

mean_results

Returns the mean of the detection and segmentation metrics over all the classes.

class_result

Returns the detection and segmentation metrics of class i.

maps

Returns the mean Average Precision (mAP) scores for IoU thresholds ranging from 0.50 to 0.95.

fitness

Returns the fitness scores, which are a single weighted combination of metrics.

ap_class_index

Returns the list of indices of classes used to compute Average Precision (AP).

results_dict

Returns the dictionary containing all the detection and segmentation metrics and fitness score.

Source code in ultralytics/utils/metrics.py
def __init__(self, save_dir=Path("."), plot=False, on_plot=None, names=()) -> None:
    """Initialize a SegmentMetrics instance with a save directory, plot flag, callback function, and class names."""
    self.save_dir = save_dir
    self.plot = plot
    self.on_plot = on_plot
    self.names = names
    self.box = Metric()
    self.seg = Metric()
    self.speed = {"preprocess": 0.0, "inference": 0.0, "loss": 0.0, "postprocess": 0.0}
    self.task = "segment"

ap_class_index property

ap_class_index

Boxes and masks have the same ap_class_index.

curves property

curves

Returns a list of curves for accessing specific metrics curves.

curves_results property

curves_results

Returns dictionary of computed performance metrics and statistics.

fitness property

fitness

Get the fitness score for both segmentation and bounding box models.

keys property

keys

Returns a list of keys for accessing metrics.

maps property

maps

Returns mAP scores for object detection and semantic segmentation models.

results_dict property

results_dict

Returns results of object detection model for evaluation.

class_result

class_result(i)

Returns classification results for a specified class index.

Source code in ultralytics/utils/metrics.py
def class_result(self, i):
    """Returns classification results for a specified class index."""
    return self.box.class_result(i) + self.seg.class_result(i)

mean_results

mean_results()

Return the mean metrics for bounding box and segmentation results.

Source code in ultralytics/utils/metrics.py
def mean_results(self):
    """Return the mean metrics for bounding box and segmentation results."""
    return self.box.mean_results() + self.seg.mean_results()

process

process(tp, tp_m, conf, pred_cls, target_cls)

Processes the detection and segmentation metrics over the given set of predictions.

Parameters:

Name Type Description Default
tp list

List of True Positive boxes.

required
tp_m list

List of True Positive masks.

required
conf list

List of confidence scores.

required
pred_cls list

List of predicted classes.

required
target_cls list

List of target classes.

required
Source code in ultralytics/utils/metrics.py
def process(self, tp, tp_m, conf, pred_cls, target_cls):
    """
    Processes the detection and segmentation metrics over the given set of predictions.

    Args:
        tp (list): List of True Positive boxes.
        tp_m (list): List of True Positive masks.
        conf (list): List of confidence scores.
        pred_cls (list): List of predicted classes.
        target_cls (list): List of target classes.
    """

    results_mask = ap_per_class(
        tp_m,
        conf,
        pred_cls,
        target_cls,
        plot=self.plot,
        on_plot=self.on_plot,
        save_dir=self.save_dir,
        names=self.names,
        prefix="Mask",
    )[2:]
    self.seg.nc = len(self.names)
    self.seg.update(results_mask)
    results_box = ap_per_class(
        tp,
        conf,
        pred_cls,
        target_cls,
        plot=self.plot,
        on_plot=self.on_plot,
        save_dir=self.save_dir,
        names=self.names,
        prefix="Box",
    )[2:]
    self.box.nc = len(self.names)
    self.box.update(results_box)





ultralytics.utils.metrics.PoseMetrics

PoseMetrics(save_dir=Path('.'), plot=False, on_plot=None, names=())

Bases: SegmentMetrics

Calculates and aggregates detection and pose metrics over a given set of classes.

Parameters:

Name Type Description Default
save_dir Path

Path to the directory where the output plots should be saved. Default is the current directory.

Path('.')
plot bool

Whether to save the detection and segmentation plots. Default is False.

False
on_plot func

An optional callback to pass plots path and data when they are rendered. Defaults to None.

None
names list

List of class names. Default is an empty list.

()

Attributes:

Name Type Description
save_dir Path

Path to the directory where the output plots should be saved.

plot bool

Whether to save the detection and segmentation plots.

on_plot func

An optional callback to pass plots path and data when they are rendered.

names list

List of class names.

box Metric

An instance of the Metric class to calculate box detection metrics.

pose Metric

An instance of the Metric class to calculate mask segmentation metrics.

speed dict

Dictionary to store the time taken in different phases of inference.

Methods:

Name Description
process

Processes metrics over the given set of predictions.

mean_results

Returns the mean of the detection and segmentation metrics over all the classes.

class_result

Returns the detection and segmentation metrics of class i.

maps

Returns the mean Average Precision (mAP) scores for IoU thresholds ranging from 0.50 to 0.95.

fitness

Returns the fitness scores, which are a single weighted combination of metrics.

ap_class_index

Returns the list of indices of classes used to compute Average Precision (AP).

results_dict

Returns the dictionary containing all the detection and segmentation metrics and fitness score.

Source code in ultralytics/utils/metrics.py
def __init__(self, save_dir=Path("."), plot=False, on_plot=None, names=()) -> None:
    """Initialize the PoseMetrics class with directory path, class names, and plotting options."""
    super().__init__(save_dir, plot, names)
    self.save_dir = save_dir
    self.plot = plot
    self.on_plot = on_plot
    self.names = names
    self.box = Metric()
    self.pose = Metric()
    self.speed = {"preprocess": 0.0, "inference": 0.0, "loss": 0.0, "postprocess": 0.0}
    self.task = "pose"

curves property

curves

Returns a list of curves for accessing specific metrics curves.

curves_results property

curves_results

Returns dictionary of computed performance metrics and statistics.

fitness property

fitness

Computes classification metrics and speed using the targets and pred inputs.

keys property

keys

Returns list of evaluation metric keys.

maps property

maps

Returns the mean average precision (mAP) per class for both box and pose detections.

class_result

class_result(i)

Return the class-wise detection results for a specific class i.

Source code in ultralytics/utils/metrics.py
def class_result(self, i):
    """Return the class-wise detection results for a specific class i."""
    return self.box.class_result(i) + self.pose.class_result(i)

mean_results

mean_results()

Return the mean results of box and pose.

Source code in ultralytics/utils/metrics.py
def mean_results(self):
    """Return the mean results of box and pose."""
    return self.box.mean_results() + self.pose.mean_results()

process

process(tp, tp_p, conf, pred_cls, target_cls)

Processes the detection and pose metrics over the given set of predictions.

Parameters:

Name Type Description Default
tp list

List of True Positive boxes.

required
tp_p list

List of True Positive keypoints.

required
conf list

List of confidence scores.

required
pred_cls list

List of predicted classes.

required
target_cls list

List of target classes.

required
Source code in ultralytics/utils/metrics.py
def process(self, tp, tp_p, conf, pred_cls, target_cls):
    """
    Processes the detection and pose metrics over the given set of predictions.

    Args:
        tp (list): List of True Positive boxes.
        tp_p (list): List of True Positive keypoints.
        conf (list): List of confidence scores.
        pred_cls (list): List of predicted classes.
        target_cls (list): List of target classes.
    """

    results_pose = ap_per_class(
        tp_p,
        conf,
        pred_cls,
        target_cls,
        plot=self.plot,
        on_plot=self.on_plot,
        save_dir=self.save_dir,
        names=self.names,
        prefix="Pose",
    )[2:]
    self.pose.nc = len(self.names)
    self.pose.update(results_pose)
    results_box = ap_per_class(
        tp,
        conf,
        pred_cls,
        target_cls,
        plot=self.plot,
        on_plot=self.on_plot,
        save_dir=self.save_dir,
        names=self.names,
        prefix="Box",
    )[2:]
    self.box.nc = len(self.names)
    self.box.update(results_box)





ultralytics.utils.metrics.ClassifyMetrics

ClassifyMetrics()

Bases: SimpleClass

Class for computing classification metrics including top-1 and top-5 accuracy.

Attributes:

Name Type Description
top1 float

The top-1 accuracy.

top5 float

The top-5 accuracy.

speed Dict[str, float]

A dictionary containing the time taken for each step in the pipeline.

fitness float

The fitness of the model, which is equal to top-5 accuracy.

results_dict Dict[str, Union[float, str]]

A dictionary containing the classification metrics and fitness.

keys List[str]

A list of keys for the results_dict.

Methods:

Name Description
process

Processes the targets and predictions to compute classification metrics.

Source code in ultralytics/utils/metrics.py
def __init__(self) -> None:
    """Initialize a ClassifyMetrics instance."""
    self.top1 = 0
    self.top5 = 0
    self.speed = {"preprocess": 0.0, "inference": 0.0, "loss": 0.0, "postprocess": 0.0}
    self.task = "classify"

curves property

curves

Returns a list of curves for accessing specific metrics curves.

curves_results property

curves_results

Returns a list of curves for accessing specific metrics curves.

fitness property

fitness

Returns mean of top-1 and top-5 accuracies as fitness score.

keys property

keys

Returns a list of keys for the results_dict property.

results_dict property

results_dict

Returns a dictionary with model's performance metrics and fitness score.

process

process(targets, pred)

Target classes and predicted classes.

Source code in ultralytics/utils/metrics.py
def process(self, targets, pred):
    """Target classes and predicted classes."""
    pred, targets = torch.cat(pred), torch.cat(targets)
    correct = (targets[:, None] == pred).float()
    acc = torch.stack((correct[:, 0], correct.max(1).values), dim=1)  # (top1, top5) accuracy
    self.top1, self.top5 = acc.mean(0).tolist()





ultralytics.utils.metrics.OBBMetrics

OBBMetrics(save_dir=Path('.'), plot=False, on_plot=None, names=())

Bases: SimpleClass

Metrics for evaluating oriented bounding box (OBB) detection, see https://arxiv.org/pdf/2106.06072.pdf.

Source code in ultralytics/utils/metrics.py
def __init__(self, save_dir=Path("."), plot=False, on_plot=None, names=()) -> None:
    """Initialize an OBBMetrics instance with directory, plotting, callback, and class names."""
    self.save_dir = save_dir
    self.plot = plot
    self.on_plot = on_plot
    self.names = names
    self.box = Metric()
    self.speed = {"preprocess": 0.0, "inference": 0.0, "loss": 0.0, "postprocess": 0.0}

ap_class_index property

ap_class_index

Returns the average precision index per class.

curves property

curves

Returns a list of curves for accessing specific metrics curves.

curves_results property

curves_results

Returns a list of curves for accessing specific metrics curves.

fitness property

fitness

Returns the fitness of box object.

keys property

keys

Returns a list of keys for accessing specific metrics.

maps property

maps

Returns mean Average Precision (mAP) scores per class.

results_dict property

results_dict

Returns dictionary of computed performance metrics and statistics.

class_result

class_result(i)

Return the result of evaluating the performance of an object detection model on a specific class.

Source code in ultralytics/utils/metrics.py
def class_result(self, i):
    """Return the result of evaluating the performance of an object detection model on a specific class."""
    return self.box.class_result(i)

mean_results

mean_results()

Calculate mean of detected objects & return precision, recall, mAP50, and mAP50-95.

Source code in ultralytics/utils/metrics.py
def mean_results(self):
    """Calculate mean of detected objects & return precision, recall, mAP50, and mAP50-95."""
    return self.box.mean_results()

process

process(tp, conf, pred_cls, target_cls)

Process predicted results for object detection and update metrics.

Source code in ultralytics/utils/metrics.py
def process(self, tp, conf, pred_cls, target_cls):
    """Process predicted results for object detection and update metrics."""
    results = ap_per_class(
        tp,
        conf,
        pred_cls,
        target_cls,
        plot=self.plot,
        save_dir=self.save_dir,
        names=self.names,
        on_plot=self.on_plot,
    )[2:]
    self.box.nc = len(self.names)
    self.box.update(results)





ultralytics.utils.metrics.bbox_ioa

bbox_ioa(box1, box2, iou=False, eps=1e-07)

Calculate the intersection over box2 area given box1 and box2. Boxes are in x1y1x2y2 format.

Parameters:

Name Type Description Default
box1 ndarray

A numpy array of shape (n, 4) representing n bounding boxes.

required
box2 ndarray

A numpy array of shape (m, 4) representing m bounding boxes.

required
iou bool

Calculate the standard IoU if True else return inter_area/box2_area.

False
eps float

A small value to avoid division by zero. Defaults to 1e-7.

1e-07

Returns:

Type Description
ndarray

A numpy array of shape (n, m) representing the intersection over box2 area.

Source code in ultralytics/utils/metrics.py
def bbox_ioa(box1, box2, iou=False, eps=1e-7):
    """
    Calculate the intersection over box2 area given box1 and box2. Boxes are in x1y1x2y2 format.

    Args:
        box1 (np.ndarray): A numpy array of shape (n, 4) representing n bounding boxes.
        box2 (np.ndarray): A numpy array of shape (m, 4) representing m bounding boxes.
        iou (bool): Calculate the standard IoU if True else return inter_area/box2_area.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-7.

    Returns:
        (np.ndarray): A numpy array of shape (n, m) representing the intersection over box2 area.
    """

    # Get the coordinates of bounding boxes
    b1_x1, b1_y1, b1_x2, b1_y2 = box1.T
    b2_x1, b2_y1, b2_x2, b2_y2 = box2.T

    # Intersection area
    inter_area = (np.minimum(b1_x2[:, None], b2_x2) - np.maximum(b1_x1[:, None], b2_x1)).clip(0) * (
        np.minimum(b1_y2[:, None], b2_y2) - np.maximum(b1_y1[:, None], b2_y1)
    ).clip(0)

    # Box2 area
    area = (b2_x2 - b2_x1) * (b2_y2 - b2_y1)
    if iou:
        box1_area = (b1_x2 - b1_x1) * (b1_y2 - b1_y1)
        area = area + box1_area[:, None] - inter_area

    # Intersection over box2 area
    return inter_area / (area + eps)





ultralytics.utils.metrics.box_iou

box_iou(box1, box2, eps=1e-07)

Calculate intersection-over-union (IoU) of boxes. Both sets of boxes are expected to be in (x1, y1, x2, y2) format. Based on https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py

Parameters:

Name Type Description Default
box1 Tensor

A tensor of shape (N, 4) representing N bounding boxes.

required
box2 Tensor

A tensor of shape (M, 4) representing M bounding boxes.

required
eps float

A small value to avoid division by zero. Defaults to 1e-7.

1e-07

Returns:

Type Description
Tensor

An NxM tensor containing the pairwise IoU values for every element in box1 and box2.

Source code in ultralytics/utils/metrics.py
def box_iou(box1, box2, eps=1e-7):
    """
    Calculate intersection-over-union (IoU) of boxes. Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
    Based on https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py

    Args:
        box1 (torch.Tensor): A tensor of shape (N, 4) representing N bounding boxes.
        box2 (torch.Tensor): A tensor of shape (M, 4) representing M bounding boxes.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-7.

    Returns:
        (torch.Tensor): An NxM tensor containing the pairwise IoU values for every element in box1 and box2.
    """

    # NOTE: Need .float() to get accurate iou values
    # inter(N,M) = (rb(N,M,2) - lt(N,M,2)).clamp(0).prod(2)
    (a1, a2), (b1, b2) = box1.float().unsqueeze(1).chunk(2, 2), box2.float().unsqueeze(0).chunk(2, 2)
    inter = (torch.min(a2, b2) - torch.max(a1, b1)).clamp_(0).prod(2)

    # IoU = inter / (area1 + area2 - inter)
    return inter / ((a2 - a1).prod(2) + (b2 - b1).prod(2) - inter + eps)





ultralytics.utils.metrics.bbox_iou

bbox_iou(box1, box2, xywh=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-07)

Calculate Intersection over Union (IoU) of box1(1, 4) to box2(n, 4).

Parameters:

Name Type Description Default
box1 Tensor

A tensor representing a single bounding box with shape (1, 4).

required
box2 Tensor

A tensor representing n bounding boxes with shape (n, 4).

required
xywh bool

If True, input boxes are in (x, y, w, h) format. If False, input boxes are in (x1, y1, x2, y2) format. Defaults to True.

True
GIoU bool

If True, calculate Generalized IoU. Defaults to False.

False
DIoU bool

If True, calculate Distance IoU. Defaults to False.

False
CIoU bool

If True, calculate Complete IoU. Defaults to False.

False
eps float

A small value to avoid division by zero. Defaults to 1e-7.

1e-07

Returns:

Type Description
Tensor

IoU, GIoU, DIoU, or CIoU values depending on the specified flags.

Source code in ultralytics/utils/metrics.py
def bbox_iou(box1, box2, xywh=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7):
    """
    Calculate Intersection over Union (IoU) of box1(1, 4) to box2(n, 4).

    Args:
        box1 (torch.Tensor): A tensor representing a single bounding box with shape (1, 4).
        box2 (torch.Tensor): A tensor representing n bounding boxes with shape (n, 4).
        xywh (bool, optional): If True, input boxes are in (x, y, w, h) format. If False, input boxes are in
                               (x1, y1, x2, y2) format. Defaults to True.
        GIoU (bool, optional): If True, calculate Generalized IoU. Defaults to False.
        DIoU (bool, optional): If True, calculate Distance IoU. Defaults to False.
        CIoU (bool, optional): If True, calculate Complete IoU. Defaults to False.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-7.

    Returns:
        (torch.Tensor): IoU, GIoU, DIoU, or CIoU values depending on the specified flags.
    """

    # Get the coordinates of bounding boxes
    if xywh:  # transform from xywh to xyxy
        (x1, y1, w1, h1), (x2, y2, w2, h2) = box1.chunk(4, -1), box2.chunk(4, -1)
        w1_, h1_, w2_, h2_ = w1 / 2, h1 / 2, w2 / 2, h2 / 2
        b1_x1, b1_x2, b1_y1, b1_y2 = x1 - w1_, x1 + w1_, y1 - h1_, y1 + h1_
        b2_x1, b2_x2, b2_y1, b2_y2 = x2 - w2_, x2 + w2_, y2 - h2_, y2 + h2_
    else:  # x1, y1, x2, y2 = box1
        b1_x1, b1_y1, b1_x2, b1_y2 = box1.chunk(4, -1)
        b2_x1, b2_y1, b2_x2, b2_y2 = box2.chunk(4, -1)
        w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps
        w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps

    # Intersection area
    inter = (b1_x2.minimum(b2_x2) - b1_x1.maximum(b2_x1)).clamp_(0) * (
        b1_y2.minimum(b2_y2) - b1_y1.maximum(b2_y1)
    ).clamp_(0)

    # Union Area
    union = w1 * h1 + w2 * h2 - inter + eps

    # IoU
    iou = inter / union
    if CIoU or DIoU or GIoU:
        cw = b1_x2.maximum(b2_x2) - b1_x1.minimum(b2_x1)  # convex (smallest enclosing box) width
        ch = b1_y2.maximum(b2_y2) - b1_y1.minimum(b2_y1)  # convex height
        if CIoU or DIoU:  # Distance or Complete IoU https://arxiv.org/abs/1911.08287v1
            c2 = cw.pow(2) + ch.pow(2) + eps  # convex diagonal squared
            rho2 = (
                (b2_x1 + b2_x2 - b1_x1 - b1_x2).pow(2) + (b2_y1 + b2_y2 - b1_y1 - b1_y2).pow(2)
            ) / 4  # center dist**2
            if CIoU:  # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47
                v = (4 / math.pi**2) * ((w2 / h2).atan() - (w1 / h1).atan()).pow(2)
                with torch.no_grad():
                    alpha = v / (v - iou + (1 + eps))
                return iou - (rho2 / c2 + v * alpha)  # CIoU
            return iou - rho2 / c2  # DIoU
        c_area = cw * ch + eps  # convex area
        return iou - (c_area - union) / c_area  # GIoU https://arxiv.org/pdf/1902.09630.pdf
    return iou  # IoU





ultralytics.utils.metrics.mask_iou

mask_iou(mask1, mask2, eps=1e-07)

Calculate masks IoU.

Parameters:

Name Type Description Default
mask1 Tensor

A tensor of shape (N, n) where N is the number of ground truth objects and n is the product of image width and height.

required
mask2 Tensor

A tensor of shape (M, n) where M is the number of predicted objects and n is the product of image width and height.

required
eps float

A small value to avoid division by zero. Defaults to 1e-7.

1e-07

Returns:

Type Description
Tensor

A tensor of shape (N, M) representing masks IoU.

Source code in ultralytics/utils/metrics.py
def mask_iou(mask1, mask2, eps=1e-7):
    """
    Calculate masks IoU.

    Args:
        mask1 (torch.Tensor): A tensor of shape (N, n) where N is the number of ground truth objects and n is the
                        product of image width and height.
        mask2 (torch.Tensor): A tensor of shape (M, n) where M is the number of predicted objects and n is the
                        product of image width and height.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-7.

    Returns:
        (torch.Tensor): A tensor of shape (N, M) representing masks IoU.
    """
    intersection = torch.matmul(mask1, mask2.T).clamp_(0)
    union = (mask1.sum(1)[:, None] + mask2.sum(1)[None]) - intersection  # (area1 + area2) - intersection
    return intersection / (union + eps)





ultralytics.utils.metrics.kpt_iou

kpt_iou(kpt1, kpt2, area, sigma, eps=1e-07)

Calculate Object Keypoint Similarity (OKS).

Parameters:

Name Type Description Default
kpt1 Tensor

A tensor of shape (N, 17, 3) representing ground truth keypoints.

required
kpt2 Tensor

A tensor of shape (M, 17, 3) representing predicted keypoints.

required
area Tensor

A tensor of shape (N,) representing areas from ground truth.

required
sigma list

A list containing 17 values representing keypoint scales.

required
eps float

A small value to avoid division by zero. Defaults to 1e-7.

1e-07

Returns:

Type Description
Tensor

A tensor of shape (N, M) representing keypoint similarities.

Source code in ultralytics/utils/metrics.py
def kpt_iou(kpt1, kpt2, area, sigma, eps=1e-7):
    """
    Calculate Object Keypoint Similarity (OKS).

    Args:
        kpt1 (torch.Tensor): A tensor of shape (N, 17, 3) representing ground truth keypoints.
        kpt2 (torch.Tensor): A tensor of shape (M, 17, 3) representing predicted keypoints.
        area (torch.Tensor): A tensor of shape (N,) representing areas from ground truth.
        sigma (list): A list containing 17 values representing keypoint scales.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-7.

    Returns:
        (torch.Tensor): A tensor of shape (N, M) representing keypoint similarities.
    """
    d = (kpt1[:, None, :, 0] - kpt2[..., 0]).pow(2) + (kpt1[:, None, :, 1] - kpt2[..., 1]).pow(2)  # (N, M, 17)
    sigma = torch.tensor(sigma, device=kpt1.device, dtype=kpt1.dtype)  # (17, )
    kpt_mask = kpt1[..., 2] != 0  # (N, 17)
    e = d / ((2 * sigma).pow(2) * (area[:, None, None] + eps) * 2)  # from cocoeval
    # e = d / ((area[None, :, None] + eps) * sigma) ** 2 / 2  # from formula
    return ((-e).exp() * kpt_mask[:, None]).sum(-1) / (kpt_mask.sum(-1)[:, None] + eps)





ultralytics.utils.metrics._get_covariance_matrix

_get_covariance_matrix(boxes)

Generating covariance matrix from obbs.

Parameters:

Name Type Description Default
boxes Tensor

A tensor of shape (N, 5) representing rotated bounding boxes, with xywhr format.

required

Returns:

Type Description
Tensor

Covariance metrixs corresponding to original rotated bounding boxes.

Source code in ultralytics/utils/metrics.py
def _get_covariance_matrix(boxes):
    """
    Generating covariance matrix from obbs.

    Args:
        boxes (torch.Tensor): A tensor of shape (N, 5) representing rotated bounding boxes, with xywhr format.

    Returns:
        (torch.Tensor): Covariance metrixs corresponding to original rotated bounding boxes.
    """
    # Gaussian bounding boxes, ignore the center points (the first two columns) because they are not needed here.
    gbbs = torch.cat((boxes[:, 2:4].pow(2) / 12, boxes[:, 4:]), dim=-1)
    a, b, c = gbbs.split(1, dim=-1)
    cos = c.cos()
    sin = c.sin()
    cos2 = cos.pow(2)
    sin2 = sin.pow(2)
    return a * cos2 + b * sin2, a * sin2 + b * cos2, (a - b) * cos * sin





ultralytics.utils.metrics.probiou

probiou(obb1, obb2, CIoU=False, eps=1e-07)

Calculate the prob IoU between oriented bounding boxes, https://arxiv.org/pdf/2106.06072v1.pdf.

Parameters:

Name Type Description Default
obb1 Tensor

A tensor of shape (N, 5) representing ground truth obbs, with xywhr format.

required
obb2 Tensor

A tensor of shape (N, 5) representing predicted obbs, with xywhr format.

required
eps float

A small value to avoid division by zero. Defaults to 1e-7.

1e-07

Returns:

Type Description
Tensor

A tensor of shape (N, ) representing obb similarities.

Source code in ultralytics/utils/metrics.py
def probiou(obb1, obb2, CIoU=False, eps=1e-7):
    """
    Calculate the prob IoU between oriented bounding boxes, https://arxiv.org/pdf/2106.06072v1.pdf.

    Args:
        obb1 (torch.Tensor): A tensor of shape (N, 5) representing ground truth obbs, with xywhr format.
        obb2 (torch.Tensor): A tensor of shape (N, 5) representing predicted obbs, with xywhr format.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-7.

    Returns:
        (torch.Tensor): A tensor of shape (N, ) representing obb similarities.
    """
    x1, y1 = obb1[..., :2].split(1, dim=-1)
    x2, y2 = obb2[..., :2].split(1, dim=-1)
    a1, b1, c1 = _get_covariance_matrix(obb1)
    a2, b2, c2 = _get_covariance_matrix(obb2)

    t1 = (
        ((a1 + a2) * (y1 - y2).pow(2) + (b1 + b2) * (x1 - x2).pow(2)) / ((a1 + a2) * (b1 + b2) - (c1 + c2).pow(2) + eps)
    ) * 0.25
    t2 = (((c1 + c2) * (x2 - x1) * (y1 - y2)) / ((a1 + a2) * (b1 + b2) - (c1 + c2).pow(2) + eps)) * 0.5
    t3 = (
        ((a1 + a2) * (b1 + b2) - (c1 + c2).pow(2))
        / (4 * ((a1 * b1 - c1.pow(2)).clamp_(0) * (a2 * b2 - c2.pow(2)).clamp_(0)).sqrt() + eps)
        + eps
    ).log() * 0.5
    bd = (t1 + t2 + t3).clamp(eps, 100.0)
    hd = (1.0 - (-bd).exp() + eps).sqrt()
    iou = 1 - hd
    if CIoU:  # only include the wh aspect ratio part
        w1, h1 = obb1[..., 2:4].split(1, dim=-1)
        w2, h2 = obb2[..., 2:4].split(1, dim=-1)
        v = (4 / math.pi**2) * ((w2 / h2).atan() - (w1 / h1).atan()).pow(2)
        with torch.no_grad():
            alpha = v / (v - iou + (1 + eps))
        return iou - v * alpha  # CIoU
    return iou





ultralytics.utils.metrics.batch_probiou

batch_probiou(obb1, obb2, eps=1e-07)

Calculate the prob IoU between oriented bounding boxes, https://arxiv.org/pdf/2106.06072v1.pdf.

Parameters:

Name Type Description Default
obb1 Tensor | ndarray

A tensor of shape (N, 5) representing ground truth obbs, with xywhr format.

required
obb2 Tensor | ndarray

A tensor of shape (M, 5) representing predicted obbs, with xywhr format.

required
eps float

A small value to avoid division by zero. Defaults to 1e-7.

1e-07

Returns:

Type Description
Tensor

A tensor of shape (N, M) representing obb similarities.

Source code in ultralytics/utils/metrics.py
def batch_probiou(obb1, obb2, eps=1e-7):
    """
    Calculate the prob IoU between oriented bounding boxes, https://arxiv.org/pdf/2106.06072v1.pdf.

    Args:
        obb1 (torch.Tensor | np.ndarray): A tensor of shape (N, 5) representing ground truth obbs, with xywhr format.
        obb2 (torch.Tensor | np.ndarray): A tensor of shape (M, 5) representing predicted obbs, with xywhr format.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-7.

    Returns:
        (torch.Tensor): A tensor of shape (N, M) representing obb similarities.
    """
    obb1 = torch.from_numpy(obb1) if isinstance(obb1, np.ndarray) else obb1
    obb2 = torch.from_numpy(obb2) if isinstance(obb2, np.ndarray) else obb2

    x1, y1 = obb1[..., :2].split(1, dim=-1)
    x2, y2 = (x.squeeze(-1)[None] for x in obb2[..., :2].split(1, dim=-1))
    a1, b1, c1 = _get_covariance_matrix(obb1)
    a2, b2, c2 = (x.squeeze(-1)[None] for x in _get_covariance_matrix(obb2))

    t1 = (
        ((a1 + a2) * (y1 - y2).pow(2) + (b1 + b2) * (x1 - x2).pow(2)) / ((a1 + a2) * (b1 + b2) - (c1 + c2).pow(2) + eps)
    ) * 0.25
    t2 = (((c1 + c2) * (x2 - x1) * (y1 - y2)) / ((a1 + a2) * (b1 + b2) - (c1 + c2).pow(2) + eps)) * 0.5
    t3 = (
        ((a1 + a2) * (b1 + b2) - (c1 + c2).pow(2))
        / (4 * ((a1 * b1 - c1.pow(2)).clamp_(0) * (a2 * b2 - c2.pow(2)).clamp_(0)).sqrt() + eps)
        + eps
    ).log() * 0.5
    bd = (t1 + t2 + t3).clamp(eps, 100.0)
    hd = (1.0 - (-bd).exp() + eps).sqrt()
    return 1 - hd





ultralytics.utils.metrics.smooth_BCE

smooth_BCE(eps=0.1)

Computes smoothed positive and negative Binary Cross-Entropy targets.

This function calculates positive and negative label smoothing BCE targets based on a given epsilon value. For implementation details, refer to https://github.com/ultralytics/yolov3/issues/238#issuecomment-598028441.

Parameters:

Name Type Description Default
eps float

The epsilon value for label smoothing. Defaults to 0.1.

0.1

Returns:

Type Description
tuple

A tuple containing the positive and negative label smoothing BCE targets.

Source code in ultralytics/utils/metrics.py
def smooth_BCE(eps=0.1):
    """
    Computes smoothed positive and negative Binary Cross-Entropy targets.

    This function calculates positive and negative label smoothing BCE targets based on a given epsilon value.
    For implementation details, refer to https://github.com/ultralytics/yolov3/issues/238#issuecomment-598028441.

    Args:
        eps (float, optional): The epsilon value for label smoothing. Defaults to 0.1.

    Returns:
        (tuple): A tuple containing the positive and negative label smoothing BCE targets.
    """
    return 1.0 - 0.5 * eps, 0.5 * eps





ultralytics.utils.metrics.smooth

smooth(y, f=0.05)

Box filter of fraction f.

Source code in ultralytics/utils/metrics.py
def smooth(y, f=0.05):
    """Box filter of fraction f."""
    nf = round(len(y) * f * 2) // 2 + 1  # number of filter elements (must be odd)
    p = np.ones(nf // 2)  # ones padding
    yp = np.concatenate((p * y[0], y, p * y[-1]), 0)  # y padded
    return np.convolve(yp, np.ones(nf) / nf, mode="valid")  # y-smoothed





ultralytics.utils.metrics.plot_pr_curve

plot_pr_curve(px, py, ap, save_dir=Path('pr_curve.png'), names=(), on_plot=None)

Plots a precision-recall curve.

Source code in ultralytics/utils/metrics.py
@plt_settings()
def plot_pr_curve(px, py, ap, save_dir=Path("pr_curve.png"), names=(), on_plot=None):
    """Plots a precision-recall curve."""
    fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True)
    py = np.stack(py, axis=1)

    if 0 < len(names) < 21:  # display per-class legend if < 21 classes
        for i, y in enumerate(py.T):
            ax.plot(px, y, linewidth=1, label=f"{names[i]} {ap[i, 0]:.3f}")  # plot(recall, precision)
    else:
        ax.plot(px, py, linewidth=1, color="grey")  # plot(recall, precision)

    ax.plot(px, py.mean(1), linewidth=3, color="blue", label="all classes %.3f mAP@0.5" % ap[:, 0].mean())
    ax.set_xlabel("Recall")
    ax.set_ylabel("Precision")
    ax.set_xlim(0, 1)
    ax.set_ylim(0, 1)
    ax.legend(bbox_to_anchor=(1.04, 1), loc="upper left")
    ax.set_title("Precision-Recall Curve")
    fig.savefig(save_dir, dpi=250)
    plt.close(fig)
    if on_plot:
        on_plot(save_dir)





ultralytics.utils.metrics.plot_mc_curve

plot_mc_curve(px, py, save_dir=Path('mc_curve.png'), names=(), xlabel='Confidence', ylabel='Metric', on_plot=None)

Plots a metric-confidence curve.

Source code in ultralytics/utils/metrics.py
@plt_settings()
def plot_mc_curve(px, py, save_dir=Path("mc_curve.png"), names=(), xlabel="Confidence", ylabel="Metric", on_plot=None):
    """Plots a metric-confidence curve."""
    fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True)

    if 0 < len(names) < 21:  # display per-class legend if < 21 classes
        for i, y in enumerate(py):
            ax.plot(px, y, linewidth=1, label=f"{names[i]}")  # plot(confidence, metric)
    else:
        ax.plot(px, py.T, linewidth=1, color="grey")  # plot(confidence, metric)

    y = smooth(py.mean(0), 0.05)
    ax.plot(px, y, linewidth=3, color="blue", label=f"all classes {y.max():.2f} at {px[y.argmax()]:.3f}")
    ax.set_xlabel(xlabel)
    ax.set_ylabel(ylabel)
    ax.set_xlim(0, 1)
    ax.set_ylim(0, 1)
    ax.legend(bbox_to_anchor=(1.04, 1), loc="upper left")
    ax.set_title(f"{ylabel}-Confidence Curve")
    fig.savefig(save_dir, dpi=250)
    plt.close(fig)
    if on_plot:
        on_plot(save_dir)





ultralytics.utils.metrics.compute_ap

compute_ap(recall, precision)

Compute the average precision (AP) given the recall and precision curves.

Parameters:

Name Type Description Default
recall list

The recall curve.

required
precision list

The precision curve.

required

Returns:

Type Description
float

Average precision.

ndarray

Precision envelope curve.

ndarray

Modified recall curve with sentinel values added at the beginning and end.

Source code in ultralytics/utils/metrics.py
def compute_ap(recall, precision):
    """
    Compute the average precision (AP) given the recall and precision curves.

    Args:
        recall (list): The recall curve.
        precision (list): The precision curve.

    Returns:
        (float): Average precision.
        (np.ndarray): Precision envelope curve.
        (np.ndarray): Modified recall curve with sentinel values added at the beginning and end.
    """

    # Append sentinel values to beginning and end
    mrec = np.concatenate(([0.0], recall, [1.0]))
    mpre = np.concatenate(([1.0], precision, [0.0]))

    # Compute the precision envelope
    mpre = np.flip(np.maximum.accumulate(np.flip(mpre)))

    # Integrate area under curve
    method = "interp"  # methods: 'continuous', 'interp'
    if method == "interp":
        x = np.linspace(0, 1, 101)  # 101-point interp (COCO)
        ap = np.trapz(np.interp(x, mrec, mpre), x)  # integrate
    else:  # 'continuous'
        i = np.where(mrec[1:] != mrec[:-1])[0]  # points where x-axis (recall) changes
        ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1])  # area under curve

    return ap, mpre, mrec





ultralytics.utils.metrics.ap_per_class

ap_per_class(tp, conf, pred_cls, target_cls, plot=False, on_plot=None, save_dir=Path(), names=(), eps=1e-16, prefix='')

Computes the average precision per class for object detection evaluation.

Parameters:

Name Type Description Default
tp ndarray

Binary array indicating whether the detection is correct (True) or not (False).

required
conf ndarray

Array of confidence scores of the detections.

required
pred_cls ndarray

Array of predicted classes of the detections.

required
target_cls ndarray

Array of true classes of the detections.

required
plot bool

Whether to plot PR curves or not. Defaults to False.

False
on_plot func

A callback to pass plots path and data when they are rendered. Defaults to None.

None
save_dir Path

Directory to save the PR curves. Defaults to an empty path.

Path()
names tuple

Tuple of class names to plot PR curves. Defaults to an empty tuple.

()
eps float

A small value to avoid division by zero. Defaults to 1e-16.

1e-16
prefix str

A prefix string for saving the plot files. Defaults to an empty string.

''

Returns:

Type Description
tuple

A tuple of six arrays and one array of unique classes, where: tp (np.ndarray): True positive counts at threshold given by max F1 metric for each class.Shape: (nc,). fp (np.ndarray): False positive counts at threshold given by max F1 metric for each class. Shape: (nc,). p (np.ndarray): Precision values at threshold given by max F1 metric for each class. Shape: (nc,). r (np.ndarray): Recall values at threshold given by max F1 metric for each class. Shape: (nc,). f1 (np.ndarray): F1-score values at threshold given by max F1 metric for each class. Shape: (nc,). ap (np.ndarray): Average precision for each class at different IoU thresholds. Shape: (nc, 10). unique_classes (np.ndarray): An array of unique classes that have data. Shape: (nc,). p_curve (np.ndarray): Precision curves for each class. Shape: (nc, 1000). r_curve (np.ndarray): Recall curves for each class. Shape: (nc, 1000). f1_curve (np.ndarray): F1-score curves for each class. Shape: (nc, 1000). x (np.ndarray): X-axis values for the curves. Shape: (1000,). prec_values: Precision values at mAP@0.5 for each class. Shape: (nc, 1000).

Source code in ultralytics/utils/metrics.py
def ap_per_class(
    tp, conf, pred_cls, target_cls, plot=False, on_plot=None, save_dir=Path(), names=(), eps=1e-16, prefix=""
):
    """
    Computes the average precision per class for object detection evaluation.

    Args:
        tp (np.ndarray): Binary array indicating whether the detection is correct (True) or not (False).
        conf (np.ndarray): Array of confidence scores of the detections.
        pred_cls (np.ndarray): Array of predicted classes of the detections.
        target_cls (np.ndarray): Array of true classes of the detections.
        plot (bool, optional): Whether to plot PR curves or not. Defaults to False.
        on_plot (func, optional): A callback to pass plots path and data when they are rendered. Defaults to None.
        save_dir (Path, optional): Directory to save the PR curves. Defaults to an empty path.
        names (tuple, optional): Tuple of class names to plot PR curves. Defaults to an empty tuple.
        eps (float, optional): A small value to avoid division by zero. Defaults to 1e-16.
        prefix (str, optional): A prefix string for saving the plot files. Defaults to an empty string.

    Returns:
        (tuple): A tuple of six arrays and one array of unique classes, where:
            tp (np.ndarray): True positive counts at threshold given by max F1 metric for each class.Shape: (nc,).
            fp (np.ndarray): False positive counts at threshold given by max F1 metric for each class. Shape: (nc,).
            p (np.ndarray): Precision values at threshold given by max F1 metric for each class. Shape: (nc,).
            r (np.ndarray): Recall values at threshold given by max F1 metric for each class. Shape: (nc,).
            f1 (np.ndarray): F1-score values at threshold given by max F1 metric for each class. Shape: (nc,).
            ap (np.ndarray): Average precision for each class at different IoU thresholds. Shape: (nc, 10).
            unique_classes (np.ndarray): An array of unique classes that have data. Shape: (nc,).
            p_curve (np.ndarray): Precision curves for each class. Shape: (nc, 1000).
            r_curve (np.ndarray): Recall curves for each class. Shape: (nc, 1000).
            f1_curve (np.ndarray): F1-score curves for each class. Shape: (nc, 1000).
            x (np.ndarray): X-axis values for the curves. Shape: (1000,).
            prec_values: Precision values at mAP@0.5 for each class. Shape: (nc, 1000).
    """

    # Sort by objectness
    i = np.argsort(-conf)
    tp, conf, pred_cls = tp[i], conf[i], pred_cls[i]

    # Find unique classes
    unique_classes, nt = np.unique(target_cls, return_counts=True)
    nc = unique_classes.shape[0]  # number of classes, number of detections

    # Create Precision-Recall curve and compute AP for each class
    x, prec_values = np.linspace(0, 1, 1000), []

    # Average precision, precision and recall curves
    ap, p_curve, r_curve = np.zeros((nc, tp.shape[1])), np.zeros((nc, 1000)), np.zeros((nc, 1000))
    for ci, c in enumerate(unique_classes):
        i = pred_cls == c
        n_l = nt[ci]  # number of labels
        n_p = i.sum()  # number of predictions
        if n_p == 0 or n_l == 0:
            continue

        # Accumulate FPs and TPs
        fpc = (1 - tp[i]).cumsum(0)
        tpc = tp[i].cumsum(0)

        # Recall
        recall = tpc / (n_l + eps)  # recall curve
        r_curve[ci] = np.interp(-x, -conf[i], recall[:, 0], left=0)  # negative x, xp because xp decreases

        # Precision
        precision = tpc / (tpc + fpc)  # precision curve
        p_curve[ci] = np.interp(-x, -conf[i], precision[:, 0], left=1)  # p at pr_score

        # AP from recall-precision curve
        for j in range(tp.shape[1]):
            ap[ci, j], mpre, mrec = compute_ap(recall[:, j], precision[:, j])
            if plot and j == 0:
                prec_values.append(np.interp(x, mrec, mpre))  # precision at mAP@0.5

    prec_values = np.array(prec_values)  # (nc, 1000)

    # Compute F1 (harmonic mean of precision and recall)
    f1_curve = 2 * p_curve * r_curve / (p_curve + r_curve + eps)
    names = [v for k, v in names.items() if k in unique_classes]  # list: only classes that have data
    names = dict(enumerate(names))  # to dict
    if plot:
        plot_pr_curve(x, prec_values, ap, save_dir / f"{prefix}PR_curve.png", names, on_plot=on_plot)
        plot_mc_curve(x, f1_curve, save_dir / f"{prefix}F1_curve.png", names, ylabel="F1", on_plot=on_plot)
        plot_mc_curve(x, p_curve, save_dir / f"{prefix}P_curve.png", names, ylabel="Precision", on_plot=on_plot)
        plot_mc_curve(x, r_curve, save_dir / f"{prefix}R_curve.png", names, ylabel="Recall", on_plot=on_plot)

    i = smooth(f1_curve.mean(0), 0.1).argmax()  # max F1 index
    p, r, f1 = p_curve[:, i], r_curve[:, i], f1_curve[:, i]  # max-F1 precision, recall, F1 values
    tp = (r * nt).round()  # true positives
    fp = (tp / (p + eps) - tp).round()  # false positives
    return tp, fp, p, r, f1, ap, unique_classes.astype(int), p_curve, r_curve, f1_curve, x, prec_values





Created 2023-11-12, Updated 2024-07-21
Authors: glenn-jocher (8), Burhan-Q (1), Laughing-q (1)