.hSddlmZddlZddlmZddlmZddlZddl Z ddl m Z m Z m Z ddlmZddlmZmZmZddlmZdd lmZmZmZdd lmZGd d eZy) ) annotationsN)Path)Any)build_dataloaderbuild_yolo_dataset converter) BaseValidator)LOGGERnmsops)check_requirements)ConfusionMatrix DetMetricsbox_iou) plot_imagesceZdZdZddfd ZddZddZddZddZddZ ddZ dd Z dd Z d d Z dd Zd!d Zd"d#dZd$dZd%dZ d& d'dZd(dZd)dZd*dZd+dZ d, d-dZxZS).DetectionValidatora~ A class extending the BaseValidator class for validation based on a detection model. This class implements validation functionality specific to object detection tasks, including metrics calculation, prediction processing, and visualization of results. Attributes: is_coco (bool): Whether the dataset is COCO. is_lvis (bool): Whether the dataset is LVIS. class_map (list[int]): Mapping from model class indices to dataset class indices. metrics (DetMetrics): Object detection metrics calculator. iouv (torch.Tensor): IoU thresholds for mAP calculation. niou (int): Number of IoU thresholds. lb (list[Any]): List for storing ground truth labels for hybrid saving. jdict (list[dict[str, Any]]): List for storing JSON detection results. stats (dict[str, list[torch.Tensor]]): Dictionary for storing statistics during validation. Examples: >>> from ultralytics.models.yolo.detect import DetectionValidator >>> args = dict(model="yolo11n.pt", data="coco8.yaml") >>> validator = DetectionValidator(args=args) >>> validator() c t|||||d|_d|_d|_d|j _tjddd|_ |jj|_ t|_ y)a Initialize detection validator with necessary variables and settings. Args: dataloader (torch.utils.data.DataLoader, optional): Dataloader to use for validation. save_dir (Path, optional): Directory to save results. args (dict[str, Any], optional): Arguments for the validator. _callbacks (list[Any], optional): List of callback functions. FNdetectg?gffffff? )super__init__is_cocois_lvis class_mapargstasktorchlinspaceiouvnumelniourmetrics)self dataloadersave_dirr _callbacks __class__s `/var/www/html/ai-service/venv/lib/python3.12/site-packages/ultralytics/models/yolo/detect/val.pyrzDetectionValidator.__init__-sg XtZ@  ! NN3b1 IIOO% !| c^|jD]W\}}t|tjs!|j |j |j j dk(||<Y|jjr|djn|djdz |d<|S)z Preprocess batch of images for YOLO validation. Args: batch (dict[str, Any]): Batch containing images and annotations. Returns: (dict[str, Any]): Preprocessed batch. cuda) non_blockingimg) items isinstancerTensortodevicetyperhalffloat)r$batchkvs r) preprocesszDetectionValidator.preprocess@sKKM VDAq!U\\*44 $++:J:Jf:T4Ua V04yy~~e ))+5 yy""!!,7yy!!!! II'';'MMH,$vi7H2I2N2N1Ot/TT ?$ r*cD|jjr9dD]4}|jj|j||j 6|j |j_|j|j_|j|j_y)z8Set final values for metrics speed and confusion matrix.)TF)r& normalizeon_plotN)rrVrXplotr&rspeedr#)r$rs r)finalize_metricsz#DetectionValidator.finalize_metricssv 99??( n %%**DMMY`d`l`l*m n!ZZ (,(=(= % $  r*c|jj|j|jj|j |jj |jjS)z Calculate and return metrics statistics. Returns: (dict[str, Any]): Dictionary containing metrics results. )r&rr)r#processr&rrVr clear_stats results_dictrbs r) get_statszDetectionValidator.get_statssP dmm$))//SWS_S_`   "||(((r*cddt|jjzz}tj|d|j |jj jg|jjz|jj jdk(r-tjd|jjd|jjr|js|jdkDrt|jjrt!|jj"D]w\}}tj||j$||jj&||jj |g|jj)|zyyyyyy) z0Print training/validation set metrics per class.z %22s%11i%11iz%11.3gallrzno labels found in z, set, can not compute metrics without labelsr@N)rNr#keysr inforT nt_per_classsum mean_resultswarningrrverboserQrRstatsrap_class_indexrB nt_per_image class_result)r$pfics r) print_resultsz DetectionValidator.print_resultss] 8c$,,2C2C.D#D D B%DLL,E,E,I,I,KjdllNgNgNijjk << $ $ ( ( *a / NN00@@lm n 99  T]]tww{s4< 1  1  11!4 11!42215  H_{] r*cZ|djddk(s|djddk(r9dtj|djd|jftiSt |d|d}d|j |d|d|jjiS)a Return correct prediction matrix. Args: preds (dict[str, torch.Tensor]): Dictionary containing prediction data with 'bboxes' and 'cls' keys. batch (dict[str, Any]): Batch dictionary containing ground truth data with 'bboxes' and 'cls' keys. Returns: (dict[str, np.ndarray]): Dictionary containing 'tp' key with correct prediction matrix of shape (N, 10) for 10 IoU levels. rortpdtyperm) rrrr"boolrmatch_predictionsrr)r$rur8rrs r)rz!DetectionValidator._process_batchs <  a A %u););A)>!)C"((E%L$6$6q$9499#ETRS SeHouX7d,,U5\5<MQQSYY[\\r*c`t|j|||j||jS)al Build YOLO Dataset. Args: img_path (str): Path to the folder containing images. mode (str): `train` mode or `val` mode, users are able to customize different augmentations for each mode. batch (int, optional): Size of batches, this is for `rect`. Returns: (Dataset): YOLO dataset. )modestride)rrrDr)r$img_pathrr8s r) build_datasetz DetectionValidator.build_datasets("$))Xudiid[_[f[fggr*c |j||d}t|||jjdd|jj|j S)a Construct and return dataloader. Args: dataset_path (str): Path to the dataset. batch_size (int): Size of each batch. Returns: (torch.utils.data.DataLoader): Dataloader for validation. rP)r8rFr{)shufflerank drop_last pin_memory)rrrworkerscompilerQ)r$ dataset_path batch_sizedatasets r)get_dataloaderz!DetectionValidator.get_dataloader"sU$$\%$P   II  ii''}}  r*ctt||d|jd|dz |j|jy)z Plot validation image samples. Args: batch (dict[str, Any]): Batch containing images and annotations. ni (int): Batch index. r val_batchz _labels.jpg)rpathsfnamerBrN)rr&rBr)r$r8nis r)plot_val_samplesz#DetectionValidator.plot_val_samples8s:  "--IbT"==**LL  r*c t|D]#\}}tj|d|z|d<%|dj}|xs|jj }|D cic].}|tj |D cgc] } | |d| c} d0} }} tj| dddddf| dddddf<t|d| |d |jd |d z |j|j ycc} wcc} }w) au Plot predicted bounding boxes on input images and save the result. Args: batch (dict[str, Any]): Batch containing images and annotations. preds (list[dict[str, torch.Tensor]]): List of predictions from the model. ni (int): Batch index. max_det (Optional[int]): Maximum number of detections to plot. rnrzrNdimrmrjr.rrz _pred.jpg)imagesrrrrBr) rr ones_likerrrhcatr xyxy2xywhrr&rBr) r$r8rurrhrrrr9rw batched_predss r)plot_predictionsz#DetectionValidator.plot_predictionsHs!' BGAt %V = AD  BQx}}.TYY..W[\RSEIIu&E!qtHW~&E1MM\ \),}X7NqRTSTRTu7U)V h2A2&<  "--IbT";;**LL  'F\s$D =D D D c 2ddlm}|tj|d|dftjd|j t j|d|djd|d jdgd  j|| y) a Save YOLO detections to a txt file in normalized coordinates in a specific format. Args: predn (dict[str, torch.Tensor]): Dictionary containing predictions with keys 'bboxes', 'conf', and 'cls'. save_conf (bool): Whether to save confidence scores. shape (tuple[int, int]): Shape of the original image (height, width). file (Path): File path to save the detections. r)Resultsr@rNrmrnr{ror)pathrBboxes)r) ultralytics.engine.resultsrrruint8rBrr unsqueezer)r$rrrfilers r)rzDetectionValidator.save_one_txtes 7 HHeAha) :**))U8_eFm.E.Eb.I5QV>> result = { ... "image_id": 42, ... "file_name": "42.jpg", ... "category_id": 18, ... "bbox": [258.15, 41.29, 348.26, 243.78], ... "score": 0.236, ... } rrmNr}rnrork)image_id file_name category_idrscore) rr isnumericintr rziptolistrUappendnamerround) r$rrrrrboxbsrrws r)rzDetectionValidator.pred_to_jsonxs$F9%&yy $ 03t9dmmE(O, ArrE c!QR%j1n$ 3::<v)=)=)?uATATAVW GAq! JJ   (!%#'>>#aΛ4QU1a[4"1a[    5s/Dc ti|dtj|d|dj|d|diS)z.Scales predictions to the original image size.rmrr|r~)r~)r scale_boxesclone)r$rrs r)rzDetectionValidator.scale_predssN  coowh%%'{# -   r*c|jdz }|jddz |jrdnd|jjdz }|j |||S)a Evaluate YOLO output in JSON format and return performance statistics. Args: stats (dict[str, Any]): Current statistics dictionary. Returns: (dict[str, Any]): Updated statistics dictionary with COCO/LVIS evaluation results. zpredictions.jsonrrzinstances_val2017.jsonlvis_v1_z.json)r&rDrrrF coco_evaluate)r$r pred_json anno_jsons r) eval_jsonzDetectionValidator.eval_jsonshMM$66 IIf  +/<<'x GXX]=^ `  !!%I>>r*c  |jjr7|js |jrt |j rt jd|d|d ||fD]}|jrJ|dt|tr|gn|}t|tr|gn|}tdddl m }m}||} | j|} t!|D]M\} } || | | |jt j} |j"j$j&Dcgc] }t)t+|j,"c}| j._| j3| j5| j7| j8d |d || dd <| j8d |d || dd <|js| j8d|d|| dd <| j8d|d|| dd <| j8d|d|| dd <P|jr|d|d<|S|Scc}w#t:$r#}t j<d|Yd}~|Sd}~wwxYw)az Evaluate COCO/LVIS metrics using faster-coco-eval library. Performs evaluation using the faster-coco-eval library to compute mAP metrics for object detection. Updates the provided stats dictionary with computed metrics including mAP50, mAP50-95, and LVIS-specific metrics if applicable. Args: stats (dict[str, Any]): Dictionary to store computed metrics and statistics. pred_json (str | Path]): Path to JSON file containing predictions in COCO format. anno_json (str | Path]): Path to JSON file containing ground truth annotations in COCO format. iou_types (str | list[str]]): IoU type(s) for evaluation. Can be single string or list of strings. Common values include "bbox", "segm", "keypoints". Defaults to "bbox". suffix (str | list[str]]): Suffix to append to metric names in stats dictionary. Should correspond to iou_types if multiple types provided. Defaults to "Box". Returns: (dict[str, Any]): Updated stats dictionary containing the computed COCO/LVIS evaluation metrics. z' Evaluating faster-coco-eval mAP using z and z...z file not foundzfaster-coco-eval>=1.6.7r)COCOCOCOeval_faster)iouType lvis_styleprint_functionAP_50zmetrics/mAP50()AP_allzmetrics/mAP50-95(APrz metrics/APr(APcz metrics/APc(APfz metrics/APf(zmetrics/mAP50-95(B)fitnessz faster-coco-eval unable to run: N)rrOrrrNrUr ris_filer1rGr faster_coco_evalr r loadResrr%rim_filesrrrparamsimgIdsevaluate accumulate summarize stats_as_dict Exceptionr)r$rrr iou_typessuffixrwr r annorriou_typerPes r)rz DetectionValidator.coco_evaluatesw6 99  DLLDLLc$**o KKB9+US\R]]`a b G"I->A99;=1#_(==;>+5i+EYK9 %/%<&&"#<=BI||I.#,Y#7YKAx)dH^d^i^iCEIOOD[D[DdDd(eqT!W\\):(eCJJ%LLNNN$MMO?B>O>OPW>XEN6!9Q<.:;ADARARS[A\E-fQil^1=>||@C@Q@QRW@X VAYq\N!<=@C@Q@QRW@X VAYq\N!<=@C@Q@QRW@X VAYq\N!<=!Y$<<',-B'CE)$ u %)f  G!A!EFF  Gs9$Ir;sB#  LL6..7JJ2aar*