|h4'ddlmZddlmZmZmZmZmZddlZddl m Z m Z ddl m Z ddlmZddlmZmZddlmZGd d e Zy) )Path)AnyDictListTupleUnionN)ClassificationDatasetbuild_dataloader) BaseValidator)LOGGER)ClassifyMetricsConfusionMatrix) plot_imagesceZdZdZddfd ZdefdZdejjddfdZ de ee fde ee ffd Z d ejde ee fddfd Zdd Zd eejeejeejfdejfd Zde eeffdZdedefdZdeeefdedej6j8j:fdZddZde ee fdeddfdZ de ee fd ejdeddfdZ!xZ"S)ClassificationValidatora A class extending the BaseValidator class for validation based on a classification model. This validator handles the validation process for classification models, including metrics calculation, confusion matrix generation, and visualization of results. Attributes: targets (List[torch.Tensor]): Ground truth class labels. pred (List[torch.Tensor]): Model predictions. metrics (ClassifyMetrics): Object to calculate and store classification metrics. names (dict): Mapping of class indices to class names. nc (int): Number of classes. confusion_matrix (ConfusionMatrix): Matrix to evaluate model performance across classes. Methods: get_desc: Return a formatted string summarizing classification metrics. init_metrics: Initialize confusion matrix, class names, and tracking containers. preprocess: Preprocess input batch by moving data to device. update_metrics: Update running metrics with model predictions and batch targets. finalize_metrics: Finalize metrics including confusion matrix and processing speed. postprocess: Extract the primary prediction from model output. get_stats: Calculate and return a dictionary of metrics. build_dataset: Create a ClassificationDataset instance for validation. get_dataloader: Build and return a data loader for classification validation. print_results: Print evaluation metrics for the classification model. plot_val_samples: Plot validation image samples with their ground truth labels. plot_predictions: Plot images with their predicted class labels. Examples: >>> from ultralytics.models.yolo.classify import ClassificationValidator >>> args = dict(model="yolo11n-cls.pt", data="imagenet10") >>> validator = ClassificationValidator(args=args) >>> validator() Notes: Torchvision classification models can also be passed to the 'model' argument, i.e. model='resnet18'. Nreturnct|||||d|_d|_d|j_t |_y)a Initialize ClassificationValidator with dataloader, save directory, and other parameters. Args: dataloader (torch.utils.data.DataLoader, optional): Dataloader to use for validation. save_dir (str | Path, optional): Directory to save results. args (dict, optional): Arguments containing model and validation configuration. _callbacks (list, optional): List of callback functions to be called during validation. Examples: >>> from ultralytics.models.yolo.classify import ClassificationValidator >>> args = dict(model="yolo11n-cls.pt", data="imagenet10") >>> validator = ClassificationValidator(args=args) >>> validator() Nclassify)super__init__targetspredargstaskr metrics)self dataloadersave_dirr _callbacks __class__s c/var/www/html/test/engine/venv/lib/python3.12/site-packages/ultralytics/models/yolo/classify/val.pyrz ClassificationValidator.__init__6s; XtZ@  # &( c ddzS)z=Return a formatted string summarizing classification metrics.z %22s%11s%11s)classestop1_acctop5_accrs r!get_descz ClassificationValidator.get_descLs#'JJJr"modelc|j|_t|j|_g|_g|_t t |jj|_y)z^Initialize confusion matrix, class names, and tracking containers for predictions and targets.)namesN) r,lenncrrrlistvaluesconfusion_matrix)rr*s r! init_metricsz$ClassificationValidator.init_metricsPsJ[[ ekk"  /d5;;;M;M;O6P Qr"batchc |dj|jd|d<|jjr|djn|dj |d<|dj|j|d<|S)zTPreprocess input batch by moving data to device and converting to appropriate dtype.imgT) non_blockingcls)todevicerhalffloat)rr3s r! preprocessz"ClassificationValidator.preprocessXsnU|t{{Fe .2iinnuU|((*%,BTBTBVe U|t{{3e  r"predsctt|jd}|jj |j ddddd|fj tjj|jj |dj tjjy)a Update running metrics with model predictions and batch targets. Args: preds (torch.Tensor): Model predictions, typically logits or probabilities for each class. batch (dict): Batch data containing images and class labels. Notes: This method appends the top-N predictions (sorted by confidence in descending order) to the prediction list for later evaluation. N is limited to the minimum of 5 and the number of classes. T) descendingNr7) minr-r,rappendargsorttypetorchint32cpur)rr=r3n5s r!update_metricsz&ClassificationValidator.update_metrics_sTZZ! $ qT:1crc6BGG TXXZ[ E%L--ekk:>>@Ar"c|jj|j|j|jj r9dD]4}|jj |j||j6|j|j_ |j|j_|j|j_y)a Finalize metrics including confusion matrix and processing speed. Notes: This method processes the accumulated predictions and targets to generate the confusion matrix, optionally plots it, and updates the metrics object with speed information. Examples: >>> validator = ClassificationValidator() >>> validator.pred = [torch.tensor([[0, 1, 2]])] # Top-3 predictions for one sample >>> validator.targets = [torch.tensor([0])] # Ground truth class >>> validator.finalize_metrics() >>> print(validator.metrics.confusion_matrix) # Access the confusion matrix )TF)r normalizeon_plotN) r1process_cls_predsrrrplotsplotrrMspeedr)rrLs r!finalize_metricsz(ClassificationValidator.finalize_metricsos // 4<<H 99??( n %%**DMMY`d`l`l*m n!ZZ  $  (,(=(= %r"c<t|ttfr|dS|S)zSExtract the primary prediction from model output if it's in a list or tuple format.r) isinstancer/tuple)rr=s r! postprocessz#ClassificationValidator.postprocesss%edE];uQxFFr"c|jj|j|j|jjS)zSCalculate and return a dictionary of metrics by processing targets and predictions.)rprocessrr results_dictr(s r! get_statsz!ClassificationValidator.get_statss. T\\4995||(((r"img_pathc\t||jd|jjS)z7Create a ClassificationDataset instance for validation.F)rootraugmentprefix)r rsplit)rr[s r! build_datasetz%ClassificationValidator.build_datasets$$(EZ^ZcZcZiZijjr" dataset_path batch_sizecj|j|}t|||jjdS)aY Build and return a data loader for classification validation. Args: dataset_path (str | Path): Path to the dataset directory. batch_size (int): Number of samples per batch. Returns: (torch.utils.data.DataLoader): DataLoader object for the classification validation dataset. )rank)rar rworkers)rrbrcdatasets r!get_dataloaderz&ClassificationValidator.get_dataloaders/$$\2TYY5F5FRPPr"cddt|jjzz}tj|d|jj |jj fzy)z6Print evaluation metrics for the classification model.z%22sz%11.3gallN)r-rkeysr infotop1top5)rpfs r! print_resultsz%ClassificationValidator.print_resultssL hT\\%6%6!77 7 B%!2!2DLL4E4EFFGr"nictjt|d|d<t||jd|dz |j |j y)a Plot validation image samples with their ground truth labels. Args: batch (Dict[str, Any]): Dictionary containing batch data with 'img' (images) and 'cls' (class labels). ni (int): Batch index used for naming the output file. Examples: >>> validator = ClassificationValidator() >>> batch = {"img": torch.rand(16, 3, 224, 224), "cls": torch.randint(0, 10, (16,))} >>> validator.plot_val_samples(batch, 0) r5 batch_idx val_batchz _labels.jpg)labelsfnamer,rMN)rFaranger-rrr,rM)rr3rrs r!plot_val_samplesz(ClassificationValidator.plot_val_samplessN#\\#eEl*;<k--IbT"==**LL  r"c t|dtjt|dtj|d}t ||j d|dz |j|jy) ae Plot images with their predicted class labels and save the visualization. Args: batch (Dict[str, Any]): Batch data containing images and other information. preds (torch.Tensor): Model predictions with shape (batch_size, num_classes). ni (int): Batch index used for naming the output file. Examples: >>> validator = ClassificationValidator() >>> batch = {"img": torch.rand(16, 3, 224, 224)} >>> preds = torch.rand(16, 10) # 16 images, 10 classes >>> validator.plot_predictions(batch, preds, 0) r5r@)dim)r5rtr7ruz _pred.jpg)rwr,rMN) dictrFrxr-argmaxrrr,rM)rr3r=rr batched_predss r!plot_predictionsz(ClassificationValidator.plot_predictionsshe ll3uU|#45 U*  --IbT";;**LL  r")NNNN)rN)#__name__ __module__ __qualname____doc__rstrr)rFnnModuler2rrr<TensorrJrRrrrrVr;rZr rarintutilsdata DataLoaderrirqryr __classcell__)r s@r!rrs$L),K#KR%((//RdRS#X4S>BELLBc3hBDB >.Gu||T%,,5Gu||I\'\!]GbgbnbnG)4U +) kck.Ck Q5s+; Q QQVQ\Q\QaQaQlQl QH  d38n # $ * d38n U\\ s W[ r"r)pathlibrtypingrrrrrrFultralytics.datar r ultralytics.engine.validatorr ultralytics.utilsr ultralytics.utils.metricsr rultralytics.utils.plottingrrr'r"r!rs000 D6$F2E mE r"