+ j(^RIHtHtHt^RIHtHtHtHtH t ^RI H t H t ]!4'd^RI Ht^RIHtHt]!4'd^RIt^RIHt^RIHt]P.!]4t]!] !R R 74!R R ] 44tR#) )AnyUnionoverload)add_end_docstringsis_torch_availableis_vision_availableloggingrequires_backends) ChunkPipelinebuild_pipeline_init_args)Image) load_image valid_imagesN)BaseModelOutput)2MODEL_FOR_ZERO_SHOT_OBJECT_DETECTION_MAPPING_NAMEST)has_image_processorcaa]tRt^toRtRtRtRtRtV3Rlt ] V3RlRl4t ] V3RlRl4t RV3R lV3R lllt R t RR lt R tRRltV3RlRltRtVtV;t#)ZeroShotObjectDetectionPipelineaN Zero shot object detection pipeline using `OwlViTForObjectDetection`. This pipeline predicts bounding boxes of objects when you provide an image and a set of `candidate_labels`. Example: ```python >>> from transformers import pipeline >>> detector = pipeline(model="google/owlvit-base-patch32", task="zero-shot-object-detection") >>> detector( ... "http://images.cocodataset.org/val2017/000000039769.jpg", ... candidate_labels=["cat", "couch"], ... ) [{'score': 0.287, 'label': 'cat', 'box': {'xmin': 324, 'ymin': 20, 'xmax': 640, 'ymax': 373}}, {'score': 0.254, 'label': 'cat', 'box': {'xmin': 1, 'ymin': 55, 'xmax': 315, 'ymax': 472}}, {'score': 0.121, 'label': 'couch', 'box': {'xmin': 4, 'ymin': 0, 'xmax': 642, 'ymax': 476}}] >>> detector( ... "https://huggingface.co/datasets/Narsil/image_dummy/raw/main/parrots.png", ... candidate_labels=["head", "bird"], ... ) [{'score': 0.119, 'label': 'bird', 'box': {'xmin': 71, 'ymin': 170, 'xmax': 410, 'ymax': 508}}] ``` 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: `"zero-shot-object-detection"`. See the list of available models on [huggingface.co/models](https://huggingface.co/models?filter=zero-shot-object-detection). FTc j<\SV`!R/VB\VR4VP\4R#)visionN)super__init__r check_model_typer)selfkwargs __class__s&,r/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/pipelines/zero_shot_object_detection.pyr(ZeroShotObjectDetectionPipeline.__init__=s, "6"$) PQc <V^8dQhRS[S[R3,RS[S[S[,,RS[RS[S[S[S[3,,/#)image Image.Imagecandidate_labelsrreturn)rstrlistrdict)format __classdict__s"r __annotate__,ZeroShotObjectDetectionPipeline.__annotate__DsL##3 -.#BES /#]`# d38n #r c R#Nr)rr#r%rs&&&,r__call__(ZeroShotObjectDetectionPipeline.__call__Cs #r c <V^8dQhRS[S[S[S[3,,RS[RS[S[S[S[S[3,,,/#)r"r#rr&)r(r)r'r)r*r+s"rr,r-Is<eed4S>2eced4PTUXZ]U]P^K_F`er c R#r/r)rr#rs&&,rr0r1Hsber c  <V^8dQhRS[S[RS[S[S[S[3,,3,RS[S[S[,,R,RS[RS[S[S[S[3,,S[S[S[S[S[3,,,,/#)r"r#r$r%Nrr&)rr'r(r)r)r*r+s"rr,r-KsVVS-d38n)==>VS /D0V V d38n T$sCx.%9 : : Vr c h<RV9dVPR4p\V\\P34'dRVRV/pMZ\V\\ 34'd<\ V4'd+\ \SV` !R\W443/VB4#Tp\SV` !V3/VBpV#)a$ Detect objects (bounding boxes & classes) in the image(s) passed as inputs. Args: image (`str`, `PIL.Image` or `list[dict[str, Any]]`): The pipeline handles three types of images: - A string containing an http url pointing to an image - A string containing a local path to an image - An image loaded in PIL directly You can use this parameter to send directly a list of images, or a dataset or a generator like so: ```python >>> from transformers import pipeline >>> detector = pipeline(model="google/owlvit-base-patch32", task="zero-shot-object-detection") >>> detector( ... [ ... { ... "image": "http://images.cocodataset.org/val2017/000000039769.jpg", ... "candidate_labels": ["cat", "couch"], ... }, ... { ... "image": "http://images.cocodataset.org/val2017/000000039769.jpg", ... "candidate_labels": ["cat", "couch"], ... }, ... ] ... ) [[{'score': 0.287, 'label': 'cat', 'box': {'xmin': 324, 'ymin': 20, 'xmax': 640, 'ymax': 373}}, {'score': 0.25, 'label': 'cat', 'box': {'xmin': 1, 'ymin': 55, 'xmax': 315, 'ymax': 472}}, {'score': 0.121, 'label': 'couch', 'box': {'xmin': 4, 'ymin': 0, 'xmax': 642, 'ymax': 476}}], [{'score': 0.287, 'label': 'cat', 'box': {'xmin': 324, 'ymin': 20, 'xmax': 640, 'ymax': 373}}, {'score': 0.254, 'label': 'cat', 'box': {'xmin': 1, 'ymin': 55, 'xmax': 315, 'ymax': 472}}, {'score': 0.121, 'label': 'couch', 'box': {'xmin': 4, 'ymin': 0, 'xmax': 642, 'ymax': 476}}]] ``` candidate_labels (`str` or `list[str]` or `list[list[str]]`): What the model should recognize in the image. threshold (`float`, *optional*, defaults to 0.1): The probability necessary to make a prediction. top_k (`int`, *optional*, defaults to None): The number of top predictions that will be returned by the pipeline. If the provided number is `None` or higher than the number of predictions available, it will default to the number of predictions. 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 lists containing prediction results, one list per input image. Each list contains dictionaries with the following keys: - **label** (`str`) -- Text query corresponding to the found object. - **score** (`float`) -- Score corresponding to the object (between 0 and 1). - **box** (`dict[str,int]`) -- Bounding box of the detected object in image's original size. It is a dictionary with `x_min`, `x_max`, `y_min`, `y_max` keys. text_queriesr#r%c32"TF wrRVRV/xK R#5i)r#r%Nr).0imglabelss& r ;ZeroShotObjectDetectionPipeline.__call__..spSoKCgs$6?Sos) pop isinstancer'r r(tuplerrr0zip)rr#r%rinputsresultsrs&&&, rr0r1Ks~ V #%zz.9  ec5;;/ 0 0u&8:JKF e} - -,u2E2E pSVW\Sop  F'"64V4r c /pRV9d VR,VR&/pRV9d VR,VR&RV9d VR,VR&V/V3#)timeout thresholdtop_kr)rrpreprocess_paramspostprocess_paramss&, r_sanitize_parameters4ZeroShotObjectDetectionPipeline._sanitize_parameterssc  +1)+< i ( & .4[.A { + f *0/ w ' "&888r c#"\VR,VR7pVR,p\V\4'dVPR4p\P !VP VP..\PR7p\V4FgwrgVPVRR7pVPVRR7p V PVP4p RV\V4^, 8HRVR V/VCV CxKi R #5i) r#)rDr%,)dtypept)return_tensorsis_last target_sizecandidate_labelN)rr>r'splittorchtensorheightwidthint32 enumerate tokenizerimage_processortorMlen) rrArDr#r%rQirR text_inputsimage_featuress &&& r preprocess*ZeroShotObjectDetectionPipeline.preprocesss6'?G<!"45 & , ,/55c: llU\\5;;$?#@ T "+,<"= A...NK!11%1MN+..tzz:N1$4 5 99{!?  !   #>sC:C<cVPR4pVPR4pVPR4pVP!R/VBpRVRVRV/VCpV#)rQrRrPr)r=model)r model_inputsrQrRrPoutputs model_outputss&& r_forward(ZeroShotObjectDetectionPipeline._forwardse"&&}5 &**+<=""9-**,|,& 5FYbdkwovw r c.pVFpVR,p\V4pVPPWRVR,R7^,pVR,P4F`pVR,V,P 4p VP VR,V,^,4p RV RVRV /p VP V 4Kb K \VRR R 7pV'dVR VpV#) rRrQ)rfrE target_sizesscoresboxesscorelabelboxcVR,#)rnr)xs&r=ZeroShotObjectDetectionPipeline.postprocess..s' r T)keyreverseN)rr[post_process_object_detectionnonzeroitem_get_bounding_boxappendsorted) rrgrErFrB model_outputrorfindexrnrpresults &&&& r postprocess+ZeroShotObjectDetectionPipeline.postprocesss)L !23E*<8L**HH$ UbHcIG!*224)%0557,,WW-=e-DQ-GH!5'5%Ev& 5*&:DI fuoGr c:<V^8dQhRRRS[S[S[3,/#)r"rpz torch.Tensorr&)r)r'int)r*r+s"rr,r-s#^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)rtolist)rrprrrrbboxs&& rrz1ZeroShotObjectDetectionPipeline._get_bounding_boxs?"%!1!1!3D D D D D    r rr/)g?N)__name__ __module__ __qualname____firstlineno____doc___load_processor_load_image_processor_load_feature_extractor_load_tokenizerrrr0rIrarhrrz__static_attributes____classdictcell__ __classcell__)rr+s@@rrrs}@O #OR ##eeVVp 9&,r r)typingrrrutilsrrrr r baser r PILr image_utilsrrrTtransformers.modeling_outputsrmodels.auto.modeling_autor get_loggerrloggerrrr rrsq''kk96=^   H %,FG[m[H[r