h1,SSKJr SSKJr SSKJr SSKJr SSKr SSK r SSK J s J r SSKJr SSKJrJrJr SSKJr SS KJrJr "S S \5rg) ) annotations) ThreadPool)Path)AnyN)DetectionValidator)LOGGER NUM_THREADSops)check_requirements)SegmentMetricsmask_iouc^\rSrSrSrSSU4SjjjrSU4SjjrSU4SjjrSSjrSU4Sjjr SU4Sjjr SU4S jjr SU4S jjr SS jr SU4S jjrSU4S jjrSU4SjjrSrU=r$)SegmentationValidatoraY A class extending the DetectionValidator class for validation based on a segmentation model. This validator handles the evaluation of segmentation models, processing both bounding box and mask predictions to compute metrics such as mAP for both detection and segmentation tasks. Attributes: plot_masks (list): List to store masks for plotting. process (callable): Function to process masks based on save_json and save_txt flags. args (namespace): Arguments for the validator. metrics (SegmentMetrics): Metrics calculator for segmentation tasks. stats (dict): Dictionary to store statistics during validation. Examples: >>> from ultralytics.models.yolo.segment import SegmentationValidator >>> args = dict(model="yolo11n-seg.pt", data="coco8-seg.yaml") >>> validator = SegmentationValidator(args=args) >>> validator() ct>[TU]XX45 SUlSURl[ 5Ulg)am Initialize SegmentationValidator and set task to 'segment', metrics to SegmentMetrics. Args: dataloader (torch.utils.data.DataLoader, optional): Dataloader to use for validation. save_dir (Path, optional): Directory to save results. args (namespace, optional): Arguments for the validator. _callbacks (list, optional): List of callback functions. Nsegment)super__init__processargstaskr metrics)self dataloadersave_dirr _callbacks __class__s ]/home/james-whalen/.local/lib/python3.13/site-packages/ultralytics/models/yolo/segment/val.pyrSegmentationValidator.__init__(s0 t@ " %' cR>[TU]U5nUSR5US'U$)z Preprocess batch of images for YOLO segmentation validation. Args: batch (dict[str, Any]): Batch containing images and annotations. Returns: (dict[str, Any]): Preprocessed batch. masks)r preprocessfloat)rbatchrs rr# SegmentationValidator.preprocess7s/"5)w--/g r c2>[TU]U5 URR(a [ S5 URR(dURR (a[ RUl g[ RUl g)z Initialize metrics and select mask processing function based on save_json flag. Args: model (torch.nn.Module): Model to validate. zfaster-coco-eval>=1.6.7N) r init_metricsr save_jsonr save_txtr process_mask_native process_maskr)rmodelrs rr("SegmentationValidator.init_metricsEsZ U# 99   8 926))2E2EI[I[s.. adaqaq r c SS-$)z5Return a formatted description of evaluation metrics.z,%22s%11s%11s%11s%11s%11s%11s%11s%11s%11s%11s) ClassImages InstanceszBox(PRmAP50 mAP50-95)zMask(Pr3r4r5)rs rget_descSegmentationValidator.get_descRs$ )  r c>[US5S:XaUSSOUSn[TU] US5nURSSVs/sHnSU-PM nn[ U5HupVUR S5nURS(aUR X%XvS US 9O`[R"S/UR [RLaUOURSSQ7[RUS RS 9US 'M U$s snf) z Post-process YOLO predictions and return output detections with proto. Args: preds (list[torch.Tensor]): Raw predictions from the model. Returns: list[dict[str, torch.Tensor]]: Processed detection predictions with masks. rNextrabboxes)shape)dtypedevicer") lenr postprocessrA enumeratepoprtorchzerosr r+uint8rC) rpredsprotoximgszipred coefficientrs rrE!SegmentationValidator.postprocessbs!$E!H  2a a#E!H- % AB01Q0 'GA((7+K$$Q' UX{N% P[[a4<<33J3J#J%PUP[P[\]\^P_a++>00 M( 1sD cH>[T U]X5nUSRSnURR(aOUSUn[ R "SUS-URS9RUSS5nXV:HR5nOUSUSU:HnU(awUSVs/sH&opR[RLaUOUS-PM( nnURSS U:wa.[R"US US S S 9SnURS 5nXSS'U$s snf)a Prepare a batch for training or inference by processing images and targets. Args: si (int): Batch index. batch (dict[str, Any]): Batch data containing images and annotations. Returns: (dict[str, Any]): Prepared batch with processed annotations. clsrr"r:)rC batch_idxrNr>NbilinearF)mode align_cornersg?)r_prepare_batchrAr overlap_maskrHarangerCviewr$rr r+F interpolategt_) rsir%prepared_batchnlr"indexs mask_sizers rrY$SegmentationValidator._prepare_batch|s /: E " ( ( + 99 ! !'N2&ELLBF5<<@EEb!QOE^**,E'N5#5#;[TU]X5nUSnURSS:XdUSRSS:Xa6[R"USRSUR 4[ S9nOo[USRS5USRS5R55nURUSXF5R5R5nURSU05 U$)a Compute correct prediction matrix for a batch based on bounding boxes and optional masks. Args: preds (dict[str, torch.Tensor]): Dictionary containing predictions with keys like 'cls' and 'masks'. batch (dict[str, Any]): Dictionary containing batch data with keys like 'cls' and 'masks'. Returns: (dict[str, np.ndarray]): A dictionary containing correct prediction matrices including 'tp_m' for mask IoU. Notes: - If `masks` is True, the function computes IoU between predicted and ground truth masks. - If `overlap` is True and `masks` is True, overlapping masks are taken into account when computing IoU. Examples: >>> preds = {"cls": torch.tensor([1, 0]), "masks": torch.rand(2, 640, 640), "bboxes": torch.rand(2, 4)} >>> batch = {"cls": torch.tensor([1, 0]), "masks": torch.rand(2, 640, 640), "bboxes": torch.rand(2, 4)} >>> correct_preds = validator._process_batch(preds, batch) rTrrBr"r:tp_m)r_process_batchrAnprIniouboolr flattenr$match_predictionscpunumpyupdate)rrKr%tpgt_clsriiours rrj$SegmentationValidator._process_batchs(W #E 1u <<?a 5<#5#5a#8A#=88U5\//2DII>dKD5>11!4eGn6L6LQ6O6U6U6WXC))%,DHHJPPRD 64.! r c>UHnUSnURSURR:a.[R"SURRS35 [ R "USURR[ RS9R5US'M [TU])XX0RRS9 g)z Plot batch predictions with masks and bounding boxes. 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. r"rz&Limiting validation plots to 'max_det=z' items.Nrh)max_det) rArrxrwarningrH as_tensorrJrprplot_predictions)rr%rKnipr"rs rr{&SegmentationValidator.plot_predictionssAgJE{{1~ 1 11!G HYHYGZZbcd/B1B1B)C5;;W[[]AgJ   r99;L;L Mr c fSSKJn U"[R"USUS4[RS9SUR [ R"USUSRS5US RS5/SS 9[ R"US [ RS9S 9RXBS 9 g)aT Save YOLO detections to a txt file in normalized coordinates in a specific format. Args: predn (torch.Tensor): Predictions in the format (x1, y1, x2, y2, conf, class). save_conf (bool): Whether to save confidence scores. shape (tuple[int, int]): Shape of the original image. file (Path): File path to save the detections. r)Resultsr:rhNr@confr<rT)dimr")pathnamesboxesr") save_conf) ultralytics.engine.resultsrrkrIrJrrHcat unsqueezerzr*)rprednrrAfilers r save_one_txt"SegmentationValidator.save_one_txts 7 HHeAha) :**))U8_eFm.E.Eb.I5QV^ SSKJm U 4Sjn[R"USS5n[ [ 5nUR X45nSSS5 [T U]!X5 [W5H#upxXR[U5*U-S'M% g!,(df  NP=f)a Save one JSON result for COCO evaluation. Args: predn (dict[str, torch.Tensor]): Predictions containing bboxes, masks, confidence scores, and classes. pbatch (dict[str, Any]): Batch dictionary containing 'imgsz', 'ori_shape', 'ratio_pad', and 'im_file'. r)encodec>T"[R"USS2SS2S4SSS95SnUSRS5US'U$)z:Encode predicted masks as RLE and append results to jdict.Nr]rJ)orderrBrcountszutf-8)rkasarraydecode)rMrlers r single_encode9SegmentationValidator.pred_to_json..single_encodesFAaDjMGLMaPCM009CMJr r")r=rr:N segmentation) faster_coco_eval.core.maskrrk transposerr mapr pred_to_jsonrFjdictrD) rrpbatchr pred_maskspoolrlesrOrrrs @rr"SegmentationValidator.pred_to_jsons 6  \\%.)<  $88M6D% U+dODA9:JJD zA~ &~ 6$% $s B B"c  >0[TU]X5ES[R"[R "US[R S9RSSS5R5R5R5USUSS90E$) z.Scales predictions to the original image size.r"rhr:r=r ori_shape ratio_pad)r) r scale_predsr scale_imagerHrzrJpermute contiguousrprq)rrrrs rr!SegmentationValidator.scale_predss g!%0 S__gekkBJJ1aQRS^^`ddflln{# -  r c>URS- nURSS- UR(aSOSURRS3- n[ TU]XUSS/S S /S 9$) z;Return COCO-style instance segmentation evaluation metrics.zpredictions.jsonrrzinstances_val2017.jsonlvis_v1_z.jsonbboxsegmBoxMask)suffix)rdatais_cocorsplitr coco_evaluate)rstats pred_json anno_jsonrs r eval_jsonSegmentationValidator.eval_jsonswMM$66 IIf  +/<<'x GXX]=^ `  w$Uy66BR\aci[j$kkr )rr)NNNN)returnNone)r%dict[str, Any]rr)r-ztorch.nn.Modulerr)rstr)rKzlist[torch.Tensor]rlist[dict[str, torch.Tensor]])r`intr%rrr)rKdict[str, torch.Tensor]r%rrzdict[str, np.ndarray])r%rrKrr|rrr) rz torch.TensorrrmrAztuple[int, int]rrrr)rrrrrr)rrrrrr)rrrr)__name__ __module__ __qualname____firstlineno____doc__rr#r(r7rErYrjr{rrrr__static_attributes__ __classcell__)rs@rrrsR( ( (  r 46<N .(;.  llr r) __future__rmultiprocessing.poolrpathlibrtypingrrqrkrHtorch.nn.functionalnn functionalr]ultralytics.models.yolo.detectrultralytics.utilsrr r ultralytics.utils.checksr ultralytics.utils.metricsr r rr6r rrs?#+ =667>pl.plr