+ QV-j&^RIHtHtHt^RIt^RIHtHtH t H t H t ^RI H t Ht] !4'd ^RIHt^RIHt]!4'd ^RIHtHtHtHt] P0!]4t]!]!RR 74!R R ] 44tR#) )AnyUnionoverloadN)add_end_docstringsis_torch_availableis_vision_availableloggingrequires_backends)Pipelinebuild_pipeline_init_args)Image) load_image)*MODEL_FOR_IMAGE_SEGMENTATION_MAPPING_NAMES-MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING_NAMES-MODEL_FOR_SEMANTIC_SEGMENTATION_MAPPING_NAMES.MODEL_FOR_UNIVERSAL_SEGMENTATION_MAPPING_NAMEST)has_image_processorcaa]tRt^toRtRtRtRtRtV3Rlt Rt ] V3RlRl4t ] V3R lR l4t V3R lV3R llt RR lt RtRRltRtVtV;t#)ImageSegmentationPipelineat Image segmentation pipeline using any `AutoModelForXXXSegmentation`. This pipeline predicts masks of objects and their classes. Example: ```python >>> from transformers import pipeline >>> segmenter = pipeline(model="facebook/detr-resnet-50-panoptic") >>> segments = segmenter("https://huggingface.co/datasets/Narsil/image_dummy/raw/main/parrots.png") >>> len(segments) 2 >>> segments[0]["label"] 'bird' >>> segments[1]["label"] 'bird' >>> type(segments[0]["mask"]) # This is a black and white mask showing where is the bird on the original image. >>> segments[0]["mask"].size (768, 512) ``` This image segmentation pipeline can currently be loaded from [`pipeline`] using the following task identifier: `"image-segmentation"`. See the list of available models on [huggingface.co/models](https://huggingface.co/models?filter=image-segmentation). FTNc <\SV`!V/VB\VR4\P!4pVP \ 4VP \4VP \4VPV4R#)visionN) super__init__r rcopyupdaterrrcheck_model_type)selfargskwargsmapping __class__s&*, z/Volumes/fast/ai/experiments/ui-tars-smoke/.venv/lib/python3.14/site-packages/transformers/pipelines/image_segmentation.pyr"ImageSegmentationPipeline.__init__Dsb $)&)$)<AACDEDEEF g&c /p/pRV9dVR,VR&VR,VR&RV9d VR,VR&RV9d VR,VR&RV9d VR,VR&RV9d VR,VR&V/V3#)subtask thresholdmask_thresholdoverlap_mask_area_thresholdtimeout)rrpreprocess_kwargspostprocess_kwargss&, r"_sanitize_parameters.ImageSegmentationPipeline._sanitize_parametersNs  ,29,= y )+1)+< i ( & .4[.A { + v %39:J3K / 0 (F 2@FGd@e < =  +1)+< i ( "&888r$c f<V^8dQhRS[S[R3,RS[RS[S[S[S[3,,/#inputsz Image.Imagerreturn)rstrrlistdict)format __classdict__s"r" __annotate__&ImageSegmentationPipeline.__annotate__`s4eeuS-%78eCeDQUVY[^V^Q_L`er$c R#Nr+rr3rs&&,r"__call__"ImageSegmentationPipeline.__call___sber$c <V^8dQhRS[S[,S[R,,RS[RS[S[S[S[S[3,,,/#r1)r6r5rr7)r8r9s"r"r:r;cs?qqtCy4 +>>q#qRVW[\`adfiai\jWkRlqr$c R#r=r+r>s&&,r"r?r@bsnqr$c <V^8dQhRS[S[RS[S[,S[R,3,RS[RS[S[S[S[3,,S[S[S[S[S[3,,,,/#r1)rr5r6rr7)r8r9s"r"r:r;es`0202CS 4 ;NNO02[^02 d38n T$sCx.%9 : :02r$c t<RV9dVPR4pVf \R4h\SV` !V3/VB#)a Perform segmentation (detect masks & classes) in the image(s) passed as inputs. Args: inputs (`str`, `list[str]`, `PIL.Image` or `list[PIL.Image]`): The pipeline handles three types of images: - A string containing an HTTP(S) link pointing to an image - A string containing a local path to an image - An image loaded in PIL directly The pipeline accepts either a single image or a batch of images. Images in a batch must all be in the same format: all as HTTP(S) links, all as local paths, or all as PIL images. subtask (`str`, *optional*): Segmentation task to be performed, choose [`semantic`, `instance` and `panoptic`] depending on model capabilities. If not set, the pipeline will attempt tp resolve in the following order: `panoptic`, `instance`, `semantic`. threshold (`float`, *optional*, defaults to 0.9): Probability threshold to filter out predicted masks. mask_threshold (`float`, *optional*, defaults to 0.5): Threshold to use when turning the predicted masks into binary values. overlap_mask_area_threshold (`float`, *optional*, defaults to 0.5): Mask overlap threshold to eliminate small, disconnected segments. timeout (`float`, *optional*, defaults to None): The maximum time in seconds to wait for fetching images from the web. If None, no timeout is set and the call may block forever. Return: If the input is a single image, will return a list of dictionaries, if the input is a list of several images, will return a list of list of dictionaries corresponding to each image. The dictionaries contain the mask, label and score (where applicable) of each detected object and contains the following keys: - **label** (`str`) -- The class label identified by the model. - **mask** (`PIL.Image`) -- A binary mask of the detected object as a Pil Image of shape (width, height) of the original image. Returns a mask filled with zeros if no object is found. - **score** (*optional* `float`) -- Optionally, when the model is capable of estimating a confidence of the "object" described by the label and the mask. imageszICannot call the image-classification pipeline without an inputs argument!)pop ValueErrorrr?)rr3rr!s&&,r"r?r@esBX v ZZ)F >hi iw1&11r$c\WR7pVPVP3.pVPPP P R8XdVf/pMRV./pVP!R RV.RR/VBpVPVP4pVPVR,RVPPPRR7R,VR&M/VPV.RR 7pVPVP4pWFR &V#) )r*OneFormerConfig task_inputsrEreturn_tensorspt max_length)paddingrMrK input_ids)rErK target_sizer+) rheightwidthmodelconfigr!__name__image_processortodtype tokenizer task_seq_len)rimager&r*rPrr3s&&&& r" preprocess$ImageSegmentationPipeline.preprocesss52 ekk23 ::   & & / /3D D''3))X%XXQWXFYYtzz*F$(NN}%$::,,99# %3%  %F= !))%)NFYYtzz*F +} r$cTVPR4pVP!R/VBpW#R&V#)rPr+)rFrS)r model_inputsrP model_outputss&& r"_forward"ImageSegmentationPipeline._forwards1"&&}5  2\2 '2m$r$c RpVR9d4\VPR4'dVPPpM9VR9d3\VPR4'dVPPpVeV!VVVVVR,R7^,p.pVR,p VR,Fp WR,8H^,p \P !V P 4P\P4RR 7p VPPPV R ,,p V R ,p VPR V R V R V /4K V#VR9d\VPR4'dVPPWR,R7^,p.pVP 4p \P!V 4pVFp W8H^,p \P !V P\P4RR 7p VPPPV ,p VPR RR V R V /4K V#\!RV R\#VP4 24h)N"post_process_panoptic_segmentation"post_process_instance_segmentationrP)r'r(r) target_sizes segmentation segments_infoidL)modelabel_idscorelabelmask"post_process_semantic_segmentation)rfzSubtask z is not supported for model >Npanoptic>Ninstance>Nsemantic)hasattrrVrdrer fromarraynumpyastypenpuint8rSrTid2labelappendrpuniquerGtype)rr`r&r'r(r)fnoutputs annotationrgsegmentrornrmlabelss&&&&&& r" postprocess%ImageSegmentationPipeline.postprocesss8 ( (WT5I5IKo-p-p%%HHB * *wt7K7KMq/r/r%%HHB >#-,G*=9   GJ">2L"?33$ 5<tzz|':':288'D3O ))227:3FG(!!7E7E64"PQ 4.!* *wt7K7KMq/r/r**MM-,HNGJ"==?LYY|,F$-4t{{288'<3G ))2259!!7D'5&$"OP  xy0LTRVR\R\M]L^_` `r$r+)NN)Ng??r)rU __module__ __qualname____firstlineno____doc___load_processor_load_image_processor_load_feature_extractor_load_tokenizerrr.rr?r\rar__static_attributes____classdictcell__ __classcell__)r!r9s@@r"rrsj!FO #O'9"ee qq0202d, ,,r$r)typingrrrrvrxutilsrrrr r baser r PILr image_utilsrmodels.auto.modeling_autorrrr get_loggerrUloggerrr+r$r"rsw''kk4(   H %,FGDDHDr$