# Copyright The Lightning team. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import logging from collections.abc import Sequence from typing import Any, Callable, List, Literal, Optional, Union import numpy as np import torch import torch.distributed as dist from torch import IntTensor, Tensor from torchmetrics.detection.helpers import _fix_empty_tensors, _input_validator from torchmetrics.metric import Metric from torchmetrics.utilities.data import _cumsum from torchmetrics.utilities.imports import _MATPLOTLIB_AVAILABLE, _PYCOCOTOOLS_AVAILABLE, _TORCHVISION_AVAILABLE from torchmetrics.utilities.plot import _AX_TYPE, _PLOT_OUT_TYPE if not _MATPLOTLIB_AVAILABLE: __doctest_skip__ = ["MeanAveragePrecision.plot"] if not _TORCHVISION_AVAILABLE or not _PYCOCOTOOLS_AVAILABLE: __doctest_skip__ = ["MeanAveragePrecision.plot", "MeanAveragePrecision"] log = logging.getLogger(__name__) def compute_area(inputs: list[Any], iou_type: Literal["bbox", "segm"] = "bbox") -> Tensor: """Compute area of input depending on the specified iou_type. Default output for empty input is :class:`~torch.Tensor` """ import pycocotools.mask as mask_utils from torchvision.ops import box_area if len(inputs) == 0: return Tensor([]) if iou_type == "bbox": return box_area(torch.stack(inputs)) if iou_type == "segm": inputs = [{"size": i[0], "counts": i[1]} for i in inputs] return torch.tensor(mask_utils.area(inputs).astype("float")) raise Exception(f"IOU type {iou_type} is not supported") def compute_iou( det: list[Any], gt: list[Any], iou_type: Literal["bbox", "segm"] = "bbox", ) -> Tensor: """Compute IOU between detections and ground-truth using the specified iou_type.""" from torchvision.ops import box_iou if iou_type == "bbox": return box_iou(torch.stack(det), torch.stack(gt)) if iou_type == "segm": return _segm_iou(det, gt) raise Exception(f"IOU type {iou_type} is not supported") class BaseMetricResults(dict): """Base metric class, that allows fields for pre-defined metrics.""" def __getattr__(self, key: str) -> Tensor: """Get a specific metric attribute.""" # Using this you get the correct error message, an AttributeError instead of a KeyError if key in self: return self[key] raise AttributeError(f"No such attribute: {key}") def __setattr__(self, key: str, value: Tensor) -> None: """Set a specific metric attribute.""" self[key] = value def __delattr__(self, key: str) -> None: """Delete a specific metric attribute.""" if key in self: del self[key] raise AttributeError(f"No such attribute: {key}") class MAPMetricResults(BaseMetricResults): """Class to wrap the final mAP results.""" __slots__ = ("classes", "map", "map_50", "map_75", "map_large", "map_medium", "map_small") class MARMetricResults(BaseMetricResults): """Class to wrap the final mAR results.""" __slots__ = ("mar_1", "mar_10", "mar_100", "mar_large", "mar_medium", "mar_small") class COCOMetricResults(BaseMetricResults): """Class to wrap the final COCO metric results including various mAP/mAR values.""" __slots__ = ( "map", "map_50", "map_75", "map_large", "map_medium", "map_per_class", "map_small", "mar_1", "mar_10", "mar_100", "mar_100_per_class", "mar_large", "mar_medium", "mar_small", ) def _segm_iou(det: list[tuple[np.ndarray, np.ndarray]], gt: list[tuple[np.ndarray, np.ndarray]]) -> Tensor: """Compute IOU between detections and ground-truths using mask-IOU. Implementation is based on pycocotools toolkit for mask_utils. Args: det: A list of detection masks as ``[(RLE_SIZE, RLE_COUNTS)]``, where ``RLE_SIZE`` is (width, height) dimension of the input and RLE_COUNTS is its RLE representation; gt: A list of ground-truth masks as ``[(RLE_SIZE, RLE_COUNTS)]``, where ``RLE_SIZE`` is (width, height) dimension of the input and RLE_COUNTS is its RLE representation; """ import pycocotools.mask as mask_utils det_coco_format = [{"size": i[0], "counts": i[1]} for i in det] gt_coco_format = [{"size": i[0], "counts": i[1]} for i in gt] return torch.tensor(mask_utils.iou(det_coco_format, gt_coco_format, [False for _ in gt])) class MeanAveragePrecision(Metric): r"""Compute the `Mean-Average-Precision (mAP) and Mean-Average-Recall (mAR)`_ for object detection predictions. .. math:: \text{mAP} = \frac{1}{n} \sum_{i=1}^{n} AP_i where :math:`AP_i` is the average precision for class :math:`i` and :math:`n` is the number of classes. The average precision is defined as the area under the precision-recall curve. If argument `class_metrics` is set to ``True``, the metric will also return the mAP/mAR per class. As input to ``forward`` and ``update`` the metric accepts the following input: - ``preds`` (:class:`~List`): A list consisting of dictionaries each containing the key-values (each dictionary corresponds to a single image). Parameters that should be provided per dict - boxes: (:class:`~torch.FloatTensor`) of shape ``(num_boxes, 4)`` containing ``num_boxes`` detection boxes of the format specified in the constructor. By default, this method expects ``(xmin, ymin, xmax, ymax)`` in absolute image coordinates. - scores: :class:`~torch.FloatTensor` of shape ``(num_boxes)`` containing detection scores for the boxes. - labels: :class:`~torch.IntTensor` of shape ``(num_boxes)`` containing 0-indexed detection classes for the boxes. - masks: :class:`~torch.bool` of shape ``(num_boxes, image_height, image_width)`` containing boolean masks. Only required when `iou_type="segm"`. - ``target`` (:class:`~List`) A list consisting of dictionaries each containing the key-values (each dictionary corresponds to a single image). Parameters that should be provided per dict: - boxes: :class:`~torch.FloatTensor` of shape ``(num_boxes, 4)`` containing ``num_boxes`` ground truth boxes of the format specified in the constructor. By default, this method expects ``(xmin, ymin, xmax, ymax)`` in absolute image coordinates. - labels: :class:`~torch.IntTensor` of shape ``(num_boxes)`` containing 0-indexed ground truth classes for the boxes. - masks: :class:`~torch.bool` of shape ``(num_boxes, image_height, image_width)`` containing boolean masks. Only required when `iou_type="segm"`. As output of ``forward`` and ``compute`` the metric returns the following output: - ``map_dict``: A dictionary containing the following key-values: - map: (:class:`~torch.Tensor`) - map_small: (:class:`~torch.Tensor`) - map_medium:(:class:`~torch.Tensor`) - map_large: (:class:`~torch.Tensor`) - mar_1: (:class:`~torch.Tensor`) - mar_10: (:class:`~torch.Tensor`) - mar_100: (:class:`~torch.Tensor`) - mar_small: (:class:`~torch.Tensor`) - mar_medium: (:class:`~torch.Tensor`) - mar_large: (:class:`~torch.Tensor`) - map_50: (:class:`~torch.Tensor`) (-1 if 0.5 not in the list of iou thresholds) - map_75: (:class:`~torch.Tensor`) (-1 if 0.75 not in the list of iou thresholds) - map_per_class: (:class:`~torch.Tensor`) (-1 if class metrics are disabled) - mar_100_per_class: (:class:`~torch.Tensor`) (-1 if class metrics are disabled) - classes (:class:`~torch.Tensor`) For an example on how to use this metric check the `torchmetrics mAP example`_. .. attention:: The ``map`` score is calculated with @[ IoU=self.iou_thresholds | area=all | max_dets=max_detection_thresholds ] **Caution:** If the initialization parameters are changed, dictionary keys for mAR can change as well. The default properties are also accessible via fields and will raise an ``AttributeError`` if not available. .. important:: This metric is following the mAP implementation of `pycocotools`_ a standard implementation for the mAP metric for object detection. .. hint:: This metric requires you to have `torchvision` version 0.8.0 or newer installed (with corresponding version 1.7.0 of torch or newer). This metric requires `pycocotools` installed when iou_type is `segm`. Please install with ``pip install torchvision`` or ``pip install torchmetrics[detection]``. Args: box_format: Input format of given boxes. Supported formats are ``[`xyxy`, `xywh`, `cxcywh`]``. iou_type: Type of input (either masks or bounding-boxes) used for computing IOU. Supported IOU types are ``["bbox", "segm"]``. If using ``"segm"``, masks should be provided (see :meth:`update`). iou_thresholds: IoU thresholds for evaluation. If set to ``None`` it corresponds to the stepped range ``[0.5,...,0.95]`` with step ``0.05``. Else provide a list of floats. rec_thresholds: Recall thresholds for evaluation. If set to ``None`` it corresponds to the stepped range ``[0,...,1]`` with step ``0.01``. Else provide a list of floats. max_detection_thresholds: Thresholds on max detections per image. If set to `None` will use thresholds ``[1, 10, 100]``. Else, please provide a list of ints. class_metrics: Option to enable per-class metrics for mAP and mAR_100. Has a performance impact. kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Raises: ModuleNotFoundError: If ``torchvision`` is not installed or version installed is lower than 0.8.0 ModuleNotFoundError: If ``iou_type`` is equal to ``segm`` and ``pycocotools`` is not installed ValueError: If ``class_metrics`` is not a boolean ValueError: If ``preds`` is not of type (:class:`~List[Dict[str, Tensor]]`) ValueError: If ``target`` is not of type ``List[Dict[str, Tensor]]`` ValueError: If ``preds`` and ``target`` are not of the same length ValueError: If any of ``preds.boxes``, ``preds.scores`` and ``preds.labels`` are not of the same length ValueError: If any of ``target.boxes`` and ``target.labels`` are not of the same length ValueError: If any box is not type float and of length 4 ValueError: If any class is not type int and of length 1 ValueError: If any score is not type float and of length 1 Example: >>> from torch import tensor >>> from torchmetrics.detection import MeanAveragePrecision >>> preds = [ ... dict( ... boxes=tensor([[258.0, 41.0, 606.0, 285.0]]), ... scores=tensor([0.536]), ... labels=tensor([0]), ... ) ... ] >>> target = [ ... dict( ... boxes=tensor([[214.0, 41.0, 562.0, 285.0]]), ... labels=tensor([0]), ... ) ... ] >>> metric = MeanAveragePrecision() >>> metric.update(preds, target) >>> from pprint import pprint >>> pprint(metric.compute()) {'classes': tensor(0, dtype=torch.int32), 'map': tensor(0.6000), 'map_50': tensor(1.), 'map_75': tensor(1.), 'map_large': tensor(0.6000), 'map_medium': tensor(-1.), 'map_per_class': tensor(-1.), 'map_small': tensor(-1.), 'mar_1': tensor(0.6000), 'mar_10': tensor(0.6000), 'mar_100': tensor(0.6000), 'mar_100_per_class': tensor(-1.), 'mar_large': tensor(0.6000), 'mar_medium': tensor(-1.), 'mar_small': tensor(-1.)} """ is_differentiable: bool = False higher_is_better: Optional[bool] = True full_state_update: bool = True plot_lower_bound: float = 0.0 plot_upper_bound: float = 1.0 detections: List[Tensor] detection_scores: List[Tensor] detection_labels: List[Tensor] groundtruths: List[Tensor] groundtruth_labels: List[Tensor] def __init__( self, box_format: str = "xyxy", iou_type: Literal["bbox", "segm"] = "bbox", iou_thresholds: Optional[list[float]] = None, rec_thresholds: Optional[list[float]] = None, max_detection_thresholds: Optional[list[int]] = None, class_metrics: bool = False, **kwargs: Any, ) -> None: super().__init__(**kwargs) if not _PYCOCOTOOLS_AVAILABLE: raise ModuleNotFoundError( "`MAP` metric requires that `pycocotools` installed." " Please install with `pip install pycocotools` or `pip install torchmetrics[detection]`" ) if not _TORCHVISION_AVAILABLE: raise ModuleNotFoundError( "`MeanAveragePrecision` metric requires that `torchvision` is installed." " Please install with `pip install torchmetrics[detection]`." ) allowed_box_formats = ("xyxy", "xywh", "cxcywh") allowed_iou_types = ("segm", "bbox") if box_format not in allowed_box_formats: raise ValueError(f"Expected argument `box_format` to be one of {allowed_box_formats} but got {box_format}") self.box_format = box_format self.iou_thresholds = iou_thresholds or torch.linspace(0.5, 0.95, round((0.95 - 0.5) / 0.05) + 1).tolist() self.rec_thresholds = rec_thresholds or torch.linspace(0.0, 1.00, round(1.00 / 0.01) + 1).tolist() max_det_threshold, _ = torch.sort(IntTensor(max_detection_thresholds or [1, 10, 100])) self.max_detection_thresholds = max_det_threshold.tolist() if iou_type not in allowed_iou_types: raise ValueError(f"Expected argument `iou_type` to be one of {allowed_iou_types} but got {iou_type}") if iou_type == "segm" and not _PYCOCOTOOLS_AVAILABLE: raise ModuleNotFoundError("When `iou_type` is set to 'segm', pycocotools need to be installed") self.iou_type = iou_type self.bbox_area_ranges = { "all": (float(0**2), float(1e5**2)), "small": (float(0**2), float(32**2)), "medium": (float(32**2), float(96**2)), "large": (float(96**2), float(1e5**2)), } if not isinstance(class_metrics, bool): raise ValueError("Expected argument `class_metrics` to be a boolean") self.class_metrics = class_metrics self.add_state("detections", default=[], dist_reduce_fx=None) self.add_state("detection_scores", default=[], dist_reduce_fx=None) self.add_state("detection_labels", default=[], dist_reduce_fx=None) self.add_state("groundtruths", default=[], dist_reduce_fx=None) self.add_state("groundtruth_labels", default=[], dist_reduce_fx=None) def update(self, preds: list[dict[str, Tensor]], target: list[dict[str, Tensor]]) -> None: """Update state with predictions and targets.""" _input_validator(preds, target, iou_type=self.iou_type) for item in preds: detections = self._get_safe_item_values(item) self.detections.append(detections) # type: ignore[arg-type] self.detection_labels.append(item["labels"]) self.detection_scores.append(item["scores"]) for item in target: groundtruths = self._get_safe_item_values(item) self.groundtruths.append(groundtruths) # type: ignore[arg-type] self.groundtruth_labels.append(item["labels"]) def _move_list_states_to_cpu(self) -> None: """Move list states to cpu to save GPU memory.""" for key in self._defaults: current_val = getattr(self, key) current_to_cpu = [] if isinstance(current_val, Sequence): for cur_v in current_val: # Cannot handle RLE as Tensor if not isinstance(cur_v, tuple): cur_v = cur_v.to("cpu") current_to_cpu.append(cur_v) setattr(self, key, current_to_cpu) def _get_safe_item_values(self, item: dict[str, Any]) -> Union[Tensor, tuple]: import pycocotools.mask as mask_utils from torchvision.ops import box_convert if self.iou_type == "bbox": boxes = _fix_empty_tensors(item["boxes"]) if boxes.numel() > 0: boxes = box_convert(boxes, in_fmt=self.box_format, out_fmt="xyxy") return boxes if self.iou_type == "segm": masks = [] for i in item["masks"].cpu().numpy(): rle = mask_utils.encode(np.asfortranarray(i)) masks.append((tuple(rle["size"]), rle["counts"])) return tuple(masks) raise Exception(f"IOU type {self.iou_type} is not supported") def _get_classes(self) -> list: """Return a list of unique classes found in ground truth and detection data.""" if len(self.detection_labels) > 0 or len(self.groundtruth_labels) > 0: return torch.cat(self.detection_labels + self.groundtruth_labels).unique().tolist() return [] def _compute_iou(self, idx: int, class_id: int, max_det: int) -> Tensor: """Compute the Intersection over Union (IoU) between bounding boxes for the given image and class. Args: idx: Image Id, equivalent to the index of supplied samples class_id: Class Id of the supplied ground truth and detection labels max_det: Maximum number of evaluated detection bounding boxes """ # if self.iou_type == "bbox": gt = self.groundtruths[idx] det = self.detections[idx] gt_label_mask = (self.groundtruth_labels[idx] == class_id).nonzero().squeeze(1) det_label_mask = (self.detection_labels[idx] == class_id).nonzero().squeeze(1) if len(gt_label_mask) == 0 or len(det_label_mask) == 0: return Tensor([]) gt = [gt[i] for i in gt_label_mask] det = [det[i] for i in det_label_mask] if len(gt) == 0 or len(det) == 0: return Tensor([]) # Sort by scores and use only max detections scores = self.detection_scores[idx] scores_filtered = scores[self.detection_labels[idx] == class_id] inds = torch.argsort(scores_filtered, descending=True) # TODO Fix (only for masks is necessary) det = [det[i] for i in inds] if len(det) > max_det: det = det[:max_det] return compute_iou(det, gt, self.iou_type).to(self.device) def __evaluate_image_gt_no_preds( self, gt: Tensor, gt_label_mask: Tensor, area_range: tuple[int, int], num_iou_thrs: int ) -> dict[str, Any]: """Evaluate images with a ground truth but no predictions.""" # GTs gt = [gt[i] for i in gt_label_mask] num_gt = len(gt) areas = compute_area(gt, iou_type=self.iou_type).to(self.device) ignore_area = (areas < area_range[0]) | (areas > area_range[1]) gt_ignore, _ = torch.sort(ignore_area.to(torch.uint8)) gt_ignore = gt_ignore.to(torch.bool) # Detections num_det = 0 det_ignore = torch.zeros((num_iou_thrs, num_det), dtype=torch.bool, device=self.device) return { "dtMatches": torch.zeros((num_iou_thrs, num_det), dtype=torch.bool, device=self.device), "gtMatches": torch.zeros((num_iou_thrs, num_gt), dtype=torch.bool, device=self.device), "dtScores": torch.zeros(num_det, dtype=torch.float32, device=self.device), "gtIgnore": gt_ignore, "dtIgnore": det_ignore, } def __evaluate_image_preds_no_gt( self, det: Tensor, idx: int, det_label_mask: Tensor, max_det: int, area_range: tuple[int, int], num_iou_thrs: int, ) -> dict[str, Any]: """Evaluate images with a prediction but no ground truth.""" # GTs num_gt = 0 gt_ignore = torch.zeros(num_gt, dtype=torch.bool, device=self.device) # Detections det = [det[i] for i in det_label_mask] scores = self.detection_scores[idx] scores_filtered = scores[det_label_mask] scores_sorted, dtind = torch.sort(scores_filtered, descending=True) det = [det[i] for i in dtind] if len(det) > max_det: det = det[:max_det] num_det = len(det) det_areas = compute_area(det, iou_type=self.iou_type).to(self.device) det_ignore_area = (det_areas < area_range[0]) | (det_areas > area_range[1]) ar = det_ignore_area.reshape((1, num_det)) det_ignore = torch.repeat_interleave(ar, num_iou_thrs, 0) return { "dtMatches": torch.zeros((num_iou_thrs, num_det), dtype=torch.bool, device=self.device), "gtMatches": torch.zeros((num_iou_thrs, num_gt), dtype=torch.bool, device=self.device), "dtScores": scores_sorted.to(self.device), "gtIgnore": gt_ignore.to(self.device), "dtIgnore": det_ignore.to(self.device), } def _evaluate_image( self, idx: int, class_id: int, area_range: tuple[int, int], max_det: int, ious: dict ) -> Optional[dict]: """Perform evaluation for single class and image. Args: idx: Image Id, equivalent to the index of supplied samples. class_id: Class Id of the supplied ground truth and detection labels. area_range: List of lower and upper bounding box area threshold. max_det: Maximum number of evaluated detection bounding boxes. ious: IoU results for image and class. """ gt = self.groundtruths[idx] det = self.detections[idx] gt_label_mask = (self.groundtruth_labels[idx] == class_id).nonzero().squeeze(1) det_label_mask = (self.detection_labels[idx] == class_id).nonzero().squeeze(1) # No Gt and No predictions --> ignore image if len(gt_label_mask) == 0 and len(det_label_mask) == 0: return None num_iou_thrs = len(self.iou_thresholds) # Some GT but no predictions if len(gt_label_mask) > 0 and len(det_label_mask) == 0: return self.__evaluate_image_gt_no_preds(gt, gt_label_mask, area_range, num_iou_thrs) # Some predictions but no GT if len(gt_label_mask) == 0 and len(det_label_mask) > 0: return self.__evaluate_image_preds_no_gt(det, idx, det_label_mask, max_det, area_range, num_iou_thrs) gt = [gt[i] for i in gt_label_mask] det = [det[i] for i in det_label_mask] if len(gt) == 0 and len(det) == 0: return None if isinstance(det, dict): det = [det] if isinstance(gt, dict): gt = [gt] areas = compute_area(gt, iou_type=self.iou_type).to(self.device) ignore_area = torch.logical_or(areas < area_range[0], areas > area_range[1]) # sort dt highest score first, sort gt ignore last ignore_area_sorted, gtind = torch.sort(ignore_area.to(torch.uint8)) # Convert to uint8 temporarily and back to bool, because "Sort currently does not support bool dtype on CUDA" ignore_area_sorted = ignore_area_sorted.to(torch.bool).to(self.device) gt = [gt[i] for i in gtind] scores = self.detection_scores[idx] scores_filtered = scores[det_label_mask] scores_sorted, dtind = torch.sort(scores_filtered, descending=True) det = [det[i] for i in dtind] if len(det) > max_det: det = det[:max_det] # load computed ious ious = ious[idx, class_id][:, gtind] if len(ious[idx, class_id]) > 0 else ious[idx, class_id] num_iou_thrs = len(self.iou_thresholds) num_gt = len(gt) num_det = len(det) gt_matches = torch.zeros((num_iou_thrs, num_gt), dtype=torch.bool, device=self.device) det_matches = torch.zeros((num_iou_thrs, num_det), dtype=torch.bool, device=self.device) gt_ignore = ignore_area_sorted det_ignore = torch.zeros((num_iou_thrs, num_det), dtype=torch.bool, device=self.device) if torch.numel(ious) > 0: for idx_iou, t in enumerate(self.iou_thresholds): for idx_det, _ in enumerate(det): m = MeanAveragePrecision._find_best_gt_match(t, gt_matches, idx_iou, gt_ignore, ious, idx_det) if m == -1: continue det_ignore[idx_iou, idx_det] = gt_ignore[m] det_matches[idx_iou, idx_det] = 1 gt_matches[idx_iou, m] = 1 # set unmatched detections outside of area range to ignore det_areas = compute_area(det, iou_type=self.iou_type).to(self.device) det_ignore_area = (det_areas < area_range[0]) | (det_areas > area_range[1]) ar = det_ignore_area.reshape((1, num_det)) det_ignore = torch.logical_or( det_ignore, torch.logical_and(det_matches == 0, torch.repeat_interleave(ar, num_iou_thrs, 0)) ) return { "dtMatches": det_matches.to(self.device), "gtMatches": gt_matches.to(self.device), "dtScores": scores_sorted.to(self.device), "gtIgnore": gt_ignore.to(self.device), "dtIgnore": det_ignore.to(self.device), } @staticmethod def _find_best_gt_match( threshold: int, gt_matches: Tensor, idx_iou: float, gt_ignore: Tensor, ious: Tensor, idx_det: int ) -> int: """Return id of best ground truth match with current detection. Args: threshold: Current threshold value. gt_matches: Tensor showing if a ground truth matches for threshold ``t`` exists. idx_iou: Id of threshold ``t``. gt_ignore: Tensor showing if ground truth should be ignored. ious: IoUs for all combinations of detection and ground truth. idx_det: Id of current detection. """ previously_matched = gt_matches[idx_iou] # type: ignore[index] # Remove previously matched or ignored gts remove_mask = previously_matched | gt_ignore gt_ious = ious[idx_det] * ~remove_mask match_idx = gt_ious.argmax().item() if gt_ious[match_idx] > threshold: # type: ignore[index] return match_idx # type: ignore[return-value] return -1 def _summarize( self, results: dict, avg_prec: bool = True, iou_threshold: Optional[float] = None, area_range: str = "all", max_dets: int = 100, ) -> Tensor: """Perform evaluation for single class and image. Args: results: Dictionary including precision, recall and scores for all combinations. avg_prec: Calculate average precision. Else calculate average recall. iou_threshold: IoU threshold. If set to ``None`` it all values are used. Else results are filtered. area_range: Bounding box area range key. max_dets: Maximum detections. """ area_inds = [i for i, k in enumerate(self.bbox_area_ranges.keys()) if k == area_range] mdet_inds = [i for i, k in enumerate(self.max_detection_thresholds) if k == max_dets] if avg_prec: # dimension of precision: [TxRxKxAxM] prec = results["precision"] # IoU if iou_threshold is not None: threshold = self.iou_thresholds.index(iou_threshold) prec = prec[threshold, :, :, area_inds, mdet_inds] else: prec = prec[:, :, :, area_inds, mdet_inds] else: # dimension of recall: [TxKxAxM] prec = results["recall"] if iou_threshold is not None: threshold = self.iou_thresholds.index(iou_threshold) prec = prec[threshold, :, :, area_inds, mdet_inds] else: prec = prec[:, :, area_inds, mdet_inds] return torch.tensor([-1.0]) if len(prec[prec > -1]) == 0 else torch.mean(prec[prec > -1]) def _calculate(self, class_ids: list) -> tuple[MAPMetricResults, MARMetricResults]: """Calculate the precision and recall for all supplied classes to calculate mAP/mAR. Args: class_ids: List of label class Ids. """ img_ids = range(len(self.groundtruths)) max_detections = self.max_detection_thresholds[-1] area_ranges = self.bbox_area_ranges.values() ious = { (idx, class_id): self._compute_iou(idx, class_id, max_detections) for idx in img_ids for class_id in class_ids } eval_imgs = [ self._evaluate_image(img_id, class_id, area, max_detections, ious) # type: ignore[arg-type] for class_id in class_ids for area in area_ranges for img_id in img_ids ] num_iou_thrs = len(self.iou_thresholds) num_rec_thrs = len(self.rec_thresholds) num_classes = len(class_ids) num_bbox_areas = len(self.bbox_area_ranges) num_max_det_thresholds = len(self.max_detection_thresholds) num_imgs = len(img_ids) precision = -torch.ones((num_iou_thrs, num_rec_thrs, num_classes, num_bbox_areas, num_max_det_thresholds)) recall = -torch.ones((num_iou_thrs, num_classes, num_bbox_areas, num_max_det_thresholds)) scores = -torch.ones((num_iou_thrs, num_rec_thrs, num_classes, num_bbox_areas, num_max_det_thresholds)) # move tensors if necessary rec_thresholds_tensor = torch.tensor(self.rec_thresholds) # retrieve E at each category, area range, and max number of detections for idx_cls, _ in enumerate(class_ids): for idx_bbox_area, _ in enumerate(self.bbox_area_ranges): for idx_max_det_thresholds, max_det in enumerate(self.max_detection_thresholds): recall, precision, scores = MeanAveragePrecision.__calculate_recall_precision_scores( recall, precision, scores, idx_cls=idx_cls, idx_bbox_area=idx_bbox_area, idx_max_det_thresholds=idx_max_det_thresholds, eval_imgs=eval_imgs, rec_thresholds=rec_thresholds_tensor, max_det=max_det, num_imgs=num_imgs, num_bbox_areas=num_bbox_areas, ) return precision, recall # type: ignore[return-value] def _summarize_results(self, precisions: Tensor, recalls: Tensor) -> tuple[MAPMetricResults, MARMetricResults]: """Summarizes the precision and recall values to calculate mAP/mAR. Args: precisions: Precision values for different thresholds recalls: Recall values for different thresholds """ results = {"precision": precisions, "recall": recalls} map_metrics = MAPMetricResults() last_max_det_threshold = self.max_detection_thresholds[-1] map_metrics.map = self._summarize(results, True, max_dets=last_max_det_threshold) if 0.5 in self.iou_thresholds: map_metrics.map_50 = self._summarize(results, True, iou_threshold=0.5, max_dets=last_max_det_threshold) else: map_metrics.map_50 = torch.tensor([-1]) if 0.75 in self.iou_thresholds: map_metrics.map_75 = self._summarize(results, True, iou_threshold=0.75, max_dets=last_max_det_threshold) else: map_metrics.map_75 = torch.tensor([-1]) map_metrics.map_small = self._summarize(results, True, area_range="small", max_dets=last_max_det_threshold) map_metrics.map_medium = self._summarize(results, True, area_range="medium", max_dets=last_max_det_threshold) map_metrics.map_large = self._summarize(results, True, area_range="large", max_dets=last_max_det_threshold) mar_metrics = MARMetricResults() for max_det in self.max_detection_thresholds: mar_metrics[f"mar_{max_det}"] = self._summarize(results, False, max_dets=max_det) mar_metrics.mar_small = self._summarize(results, False, area_range="small", max_dets=last_max_det_threshold) mar_metrics.mar_medium = self._summarize(results, False, area_range="medium", max_dets=last_max_det_threshold) mar_metrics.mar_large = self._summarize(results, False, area_range="large", max_dets=last_max_det_threshold) return map_metrics, mar_metrics @staticmethod def __calculate_recall_precision_scores( recall: Tensor, precision: Tensor, scores: Tensor, idx_cls: int, idx_bbox_area: int, idx_max_det_thresholds: int, eval_imgs: list, rec_thresholds: Tensor, max_det: int, num_imgs: int, num_bbox_areas: int, ) -> tuple[Tensor, Tensor, Tensor]: num_rec_thrs = len(rec_thresholds) idx_cls_pointer = idx_cls * num_bbox_areas * num_imgs idx_bbox_area_pointer = idx_bbox_area * num_imgs # Load all image evals for current class_id and area_range img_eval_cls_bbox = [eval_imgs[idx_cls_pointer + idx_bbox_area_pointer + i] for i in range(num_imgs)] img_eval_cls_bbox = [e for e in img_eval_cls_bbox if e is not None] if not img_eval_cls_bbox: return recall, precision, scores det_scores = torch.cat([e["dtScores"][:max_det] for e in img_eval_cls_bbox]) # different sorting method generates slightly different results. # mergesort is used to be consistent as Matlab implementation. # Sort in PyTorch does not support bool types on CUDA (yet, 1.11.0) dtype = torch.uint8 if det_scores.is_cuda and det_scores.dtype is torch.bool else det_scores.dtype # Explicitly cast to uint8 to avoid error for bool inputs on CUDA to argsort inds = torch.argsort(det_scores.to(dtype), descending=True) det_scores_sorted = det_scores[inds] det_matches = torch.cat([e["dtMatches"][:, :max_det] for e in img_eval_cls_bbox], axis=1)[:, inds] # type: ignore[call-overload] det_ignore = torch.cat([e["dtIgnore"][:, :max_det] for e in img_eval_cls_bbox], axis=1)[:, inds] # type: ignore[call-overload] gt_ignore = torch.cat([e["gtIgnore"] for e in img_eval_cls_bbox]) npig = torch.count_nonzero(gt_ignore == False) # noqa: E712 if npig == 0: return recall, precision, scores tps = torch.logical_and(det_matches, torch.logical_not(det_ignore)) fps = torch.logical_and(torch.logical_not(det_matches), torch.logical_not(det_ignore)) tp_sum = _cumsum(tps, dim=1, dtype=torch.float) fp_sum = _cumsum(fps, dim=1, dtype=torch.float) for idx, (tp, fp) in enumerate(zip(tp_sum, fp_sum)): tp_len = len(tp) rc = tp / npig pr = tp / (fp + tp + torch.finfo(torch.float64).eps) prec = torch.zeros((num_rec_thrs,)) score = torch.zeros((num_rec_thrs,)) recall[idx, idx_cls, idx_bbox_area, idx_max_det_thresholds] = rc[-1] if tp_len else 0 # Remove zigzags for AUC diff_zero = torch.zeros((1,), device=pr.device) diff = torch.ones((1,), device=pr.device) while not torch.all(diff == 0): diff = torch.clamp(torch.cat(((pr[1:] - pr[:-1]), diff_zero), 0), min=0) pr += diff inds = torch.searchsorted(rc, rec_thresholds.to(rc.device), right=False) num_inds = inds.argmax() if inds.max() >= tp_len else num_rec_thrs inds = inds[:num_inds] prec[:num_inds] = pr[inds] score[:num_inds] = det_scores_sorted[inds] precision[idx, :, idx_cls, idx_bbox_area, idx_max_det_thresholds] = prec scores[idx, :, idx_cls, idx_bbox_area, idx_max_det_thresholds] = score return recall, precision, scores def compute(self) -> dict: """Compute metric.""" classes = self._get_classes() precisions, recalls = self._calculate(classes) map_val, mar_val = self._summarize_results(precisions, recalls) # type: ignore[arg-type] # if class mode is enabled, evaluate metrics per class map_per_class_values: Tensor = torch.tensor([-1.0]) mar_max_dets_per_class_values: Tensor = torch.tensor([-1.0]) if self.class_metrics: map_per_class_list = [] mar_max_dets_per_class_list = [] for class_idx, _ in enumerate(classes): cls_precisions = precisions[:, :, class_idx].unsqueeze(dim=2) cls_recalls = recalls[:, class_idx].unsqueeze(dim=1) cls_map, cls_mar = self._summarize_results(cls_precisions, cls_recalls) map_per_class_list.append(cls_map.map) mar_max_dets_per_class_list.append(cls_mar[f"mar_{self.max_detection_thresholds[-1]}"]) map_per_class_values = torch.tensor(map_per_class_list, dtype=torch.float) mar_max_dets_per_class_values = torch.tensor(mar_max_dets_per_class_list, dtype=torch.float) metrics = COCOMetricResults() metrics.update(map_val) metrics.update(mar_val) metrics.map_per_class = map_per_class_values metrics[f"mar_{self.max_detection_thresholds[-1]}_per_class"] = mar_max_dets_per_class_values metrics.classes = torch.tensor(classes, dtype=torch.int) return metrics def _apply(self, fn: Callable) -> torch.nn.Module: # type: ignore[override] """Custom apply function. Excludes the detections and groundtruths from the casting when the iou_type is set to `segm` as the state is no longer a tensor but a tuple. """ if self.iou_type == "segm": this = super()._apply(fn, exclude_state=("detections", "groundtruths")) else: this = super()._apply(fn) return this def _sync_dist(self, dist_sync_fn: Optional[Callable] = None, process_group: Optional[Any] = None) -> None: """Custom sync function. For the iou_type `segm` the detections and groundtruths are no longer tensors but tuples. Therefore, we need to gather the list of tuples and then convert it back to a list of tuples. """ super()._sync_dist(dist_sync_fn=dist_sync_fn, process_group=process_group) # type: ignore[arg-type] if self.iou_type == "segm": self.detections = self._gather_tuple_list(self.detections, process_group) # type: ignore[arg-type] self.groundtruths = self._gather_tuple_list(self.groundtruths, process_group) # type: ignore[arg-type] @staticmethod def _gather_tuple_list( list_to_gather: list[Union[tuple, Tensor]], process_group: Optional[Any] = None ) -> list[Any]: """Gather a list of tuples over multiple devices.""" world_size = dist.get_world_size(group=process_group) dist.barrier(group=process_group) list_gathered = [None for _ in range(world_size)] dist.all_gather_object(list_gathered, list_to_gather, group=process_group) return [list_gathered[rank][idx] for idx in range(len(list_gathered[0])) for rank in range(world_size)] # type: ignore[arg-type,index] def plot( self, val: Optional[Union[dict[str, Tensor], Sequence[dict[str, Tensor]]]] = None, ax: Optional[_AX_TYPE] = None ) -> _PLOT_OUT_TYPE: """Plot a single or multiple values from the metric. Args: val: Either a single result from calling `metric.forward` or `metric.compute` or a list of these results. If no value is provided, will automatically call `metric.compute` and plot that result. ax: An matplotlib axis object. If provided will add plot to that axis Returns: Figure object and Axes object Raises: ModuleNotFoundError: If `matplotlib` is not installed .. plot:: :scale: 75 >>> from torch import tensor >>> from torchmetrics.detection.mean_ap import MeanAveragePrecision >>> preds = [dict( ... boxes=tensor([[258.0, 41.0, 606.0, 285.0]]), ... scores=tensor([0.536]), ... labels=tensor([0]), ... )] >>> target = [dict( ... boxes=tensor([[214.0, 41.0, 562.0, 285.0]]), ... labels=tensor([0]), ... )] >>> metric = MeanAveragePrecision() >>> metric.update(preds, target) >>> fig_, ax_ = metric.plot() .. plot:: :scale: 75 >>> # Example plotting multiple values >>> import torch >>> from torchmetrics.detection.mean_ap import MeanAveragePrecision >>> preds = lambda: [dict( ... boxes=torch.tensor([[258.0, 41.0, 606.0, 285.0]]) + torch.randint(10, (1,4)), ... scores=torch.tensor([0.536]) + 0.1*torch.rand(1), ... labels=torch.tensor([0]), ... )] >>> target = [dict( ... boxes=torch.tensor([[214.0, 41.0, 562.0, 285.0]]), ... labels=torch.tensor([0]), ... )] >>> metric = MeanAveragePrecision() >>> vals = [] >>> for _ in range(20): ... vals.append(metric(preds(), target)) >>> fig_, ax_ = metric.plot(vals) """ return self._plot(val, ax)