+ QV-j ^RIHtHtHtHt^RIHtHtHtH t H t ^RI H t H t ]!4'd^RIHt]!4'd ^RIt^RIHtHt]'d^RIHt] P,!]4t]!] !RR 74!R R ] 44tR#) ) TYPE_CHECKINGAnyUnionoverload)add_end_docstringsis_torch_availableis_vision_availableloggingrequires_backends)Pipelinebuild_pipeline_init_args) load_imageN)(MODEL_FOR_OBJECT_DETECTION_MAPPING_NAMES,MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING_NAMES)ImageT)has_image_processorcaa]tRt^toRtRtRtRtRtV3Rlt Rt ] V3RlRl4t ] V3R lR l4t V3R lV3R llt RR lt RtRRltV3RlRltRtVtV;t#)ObjectDetectionPipelinea Object detection pipeline using any `AutoModelForObjectDetection`. This pipeline predicts bounding boxes of objects and their classes. Example: ```python >>> from transformers import pipeline >>> detector = pipeline(model="facebook/detr-resnet-50") >>> detector("https://huggingface.co/datasets/Narsil/image_dummy/raw/main/parrots.png") [{'score': 0.997, 'label': 'bird', 'box': {'xmin': 69, 'ymin': 171, 'xmax': 396, 'ymax': 507}}, {'score': 0.999, 'label': 'bird', 'box': {'xmin': 398, 'ymin': 105, 'xmax': 767, 'ymax': 507}}] >>> # x, y are expressed relative to the top left hand corner. ``` Learn more about the basics of using a pipeline in the [pipeline tutorial](../pipeline_tutorial) This object detection pipeline can currently be loaded from [`pipeline`] using the following task identifier: `"object-detection"`. See the list of available models on [huggingface.co/models](https://huggingface.co/models?filter=object-detection). FTNc<\SV`!V/VB\VR4\P!4pVP \ 4VPV4R#)visionN)super__init__r rcopyupdatercheck_model_type)selfargskwargsmapping __class__s&*, x/Volumes/fast/ai/experiments/ui-tars-smoke/.venv/lib/python3.14/site-packages/transformers/pipelines/object_detection.pyr ObjectDetectionPipeline.__init__8sH $)&)$):??ACD g&c `/pRV9d VR,VR&/pRV9d VR,VR&V/V3#)timeout threshold)rrpreprocess_paramspostprocess_kwargss&, r!_sanitize_parameters,ObjectDetectionPipeline._sanitize_parameters@sL  +1)+< i ( & .4[.A { + "&888r#c l<V^8dQhRS[S[R3,RS[RS[RS[S[S[S[3,,/#imagez Image.Imagerrreturn)rstrrlistdict)format __classdict__s"r! __annotate__$ObjectDetectionPipeline.__annotate__Js:ppeC$67ppspW[\`adfiai\jWkpr#cR#Nr'rr/rrs&&*,r!__call__ ObjectDetectionPipeline.__call__Ismpr#c <V^8dQhRS[S[,S[R,,RS[RS[RS[S[S[S[S[3,,,/#r-)r2r1rr3)r4r5s"r!r6r7MsM))#Ym!44)=@)LO) d4S>" #)r#cR#r9r'r:s&&*,r!r;r<Ls&)r#c<V^8dQhRS[S[S[S[3,,S[S[S[S[S[3,,,,/#)r.r0)r2r3r1r)r4r5s"r!r6r7Qs;"1"14S#X+?$tDQTVYQYNG[B\+\"1r#c h<RV9dRV9dVPR4VR&\SV`!V/VB#)a Detect objects (bounding boxes & 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. threshold (`float`, *optional*, defaults to 0.5): The probability necessary to make a prediction. 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: A list of dictionaries or a list of list of dictionaries containing the result. 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 following keys: - **label** (`str`) -- The class label identified by the model. - **score** (`float`) -- The score attributed by the model for that label. - **box** (`list[dict[str, int]]`) -- The bounding box of detected object in image's original size. imagesinputs)poprr;)rrrr s&*,r!r;r<Qs=@ v (&"8%zz(3F8 w000r#c>\WR7p\P!VPVP..4pVP V.RR7pVP VP4pVPe#VPVR,VR,RR7pW4R&V#))r%pt)rAreturn_tensorswordsboxes)textrHrF target_size) rtorch IntTensorheightwidthimage_processortodtype tokenizer)rr/r%rJrBs&&& r! preprocess"ObjectDetectionPipeline.preprocessus52oo ekk'B&CD %%eWT%J4::& >> %^^w`d^eF +} r#cVPR4pVP!R/VBpVPRV/VC4pVPe VR,VR&V#)rJbboxr')rCmodelr rR)r model_inputsrJoutputs model_outputss&& r!_forward ObjectDetectionPipeline._forwards\"&&}5 **,|,))=+*Q*QR >> %$0$8M& !r#c aaaVR,pSPEeV^,P4wooVVV3RlpVR,P^4PRR7P RR7wrVVP4Uu.uF*pSP P PV,NK, ppVR,P^4U u.uF q!V 4NK p p .R Op \VP4W4U u.uF&q^,V8gK\\W44NK( p p V #SPPWV4pV^,pVR,pVR,pVR,p VP4VR&VUu.uF8pSP P PVP4,NK: upVR&V Uu.uFpSPV4NK upVR&.R Op \VR,VR,VR,4U u.uFp \\W44NK p p V #uupiuup iuup iuupiuupiuup i) rJc <SP\P!SV^,,R, SV^,,R, SV^,,R, SV^,,R, .44#)ri)_get_bounding_boxrKTensor)rVrMrrNs&r! unnormalize8ObjectDetectionPipeline.postprocess..unnormalizesi--LL"T!W_t3#d1g-4"T!W_t3#d1g-4   r#logits)dimrVscoreslabelsrH)scorelabelbox)rRtolistsqueezesoftmaxmaxrWconfigid2labelzipr3rOpost_process_object_detectionitemr_)rrZr&rJrareclasses predictionrfrVrHkeysvals annotationraw_annotationsraw_annotationrirjrMrNsf&& @@r! postprocess#ObjectDetectionPipeline.postprocesss4#M2 >> %(N113MFE ,H5==a@HHRHPTTY[T\OFOV~~O_`O_djj''00<IIIc:<V^8dQhRRRS[S[S[3,/#)r.rjz torch.Tensorr0)r3r1int)r4r5s"r!r6r7s#^S#Xr#c \VP4P4wr#rERVRVRVRV/pV#)z Turns list [xmin, xmax, ymin, ymax] into dict { "xmin": xmin, ... } Args: box (`torch.Tensor`): Tensor containing the coordinates in corners format. Returns: bbox (`dict[str, int]`): Dict containing the coordinates in corners format. xminyminxmaxymax)r~rk)rrjrrrrrVs&& r!r_)ObjectDetectionPipeline._get_bounding_boxs?"%!1!1!3D D D D D    r#r'r9)g?)__name__ __module__ __qualname____firstlineno____doc___load_processor_load_image_processor_load_feature_extractor_load_tokenizerrr*rr;rSr[r{r___static_attributes____classdictcell__ __classcell__)r r5s@@r!rrsx0O #O'9pp ))"1"1H+Zr#r)typingrrrrutilsrrr r r baser r image_utilsrrKmodels.auto.modeling_autorrPILr get_loggerrloggerrr'r#r!rsw66kk4(    H %,FGkhkHkr#