o Zh @sddlmZmZmZmZddlmZmZmZm Z m Z ddl m Z m Z er+ddlmZer:ddlZddlmZmZe eZeeefZeeZee d d Gd d d e ZdS) )AnyDictListUnion)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_NAMEST)Zhas_image_processorcs|eZdZdZfddZddZdeeee fffdd Z dd d Z d d Z dddZ dddeeeffddZZS)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). csXtj|i||jdkrtd|jdt|dt}|t | |dS)NtfzThe z is only available in PyTorch.Zvision) super__init__ framework ValueError __class__r rcopyupdaterZcheck_model_type)selfargskwargsmappingrV/var/www/auris/lib/python3.10/site-packages/transformers/pipelines/object_detection.pyr4s   z ObjectDetectionPipeline.__init__cKs:i}d|vr |d|d<i}d|vr|d|d<|i|fS)Ntimeout thresholdr )rrZpreprocess_paramsZpostprocess_kwargsr r r!_sanitize_parameters?s   z,ObjectDetectionPipeline._sanitize_parametersreturncs0d|vrd|vr|d|d<tj|i|S)ai 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)popr__call__)rrrrr r!r)Hs z ObjectDetectionPipeline.__call__NcCstt||d}t|j|jgg}|j|gdd}|jdkr#||j}|j dur4|j |d|ddd}||d<|S)N)r"pt)r&return_tensorswordsboxes)textr-r+ target_size) rtorchZ IntTensorheightwidthimage_processorrtoZ torch_dtype tokenizer)rimager"r/r'r r r! preprocessls    z"ObjectDetectionPipeline.preprocesscCsF|d}|jdi|}|d|i|}|jdur!|d|d<|S)Nr/bboxr )r(modelrr5)rZ model_inputsr/Zoutputs model_outputsr r r!_forwardws   z ObjectDetectionPipeline._forward?c sL|d}jdurY|d\fdd|ddjddjdd\}}fdd |D}fd d |d dD}gd fd d t|||D}|Sj||} | d} | d}| d}| d}|| d<fdd |D| d<fdd |D| d<gd fdd t| d| d| dD}|S)Nr/rc sHt|dd|dd|dd|ddgS)Nrir r)_get_bounding_boxr0ZTensor)r8)r1rr2r r! unnormalizesz8ObjectDetectionPipeline.postprocess..unnormalizeZlogits)dimcsg|] }jjj|qSr )r9configid2label).0Z predictionrr r! sz7ObjectDetectionPipeline.postprocess..csg|]}|qSr r )rDr8)r?r r!rFsr8)Zscorelabelboxcs&g|]}|dkrtt|qS)rdictziprDvals)keysr#r r!rFs&scoreslabelsr-csg|] }jjj|qSr )r9rBrCitem)rDrGrEr r!rFscsg|]}|qSr )r>)rDrHrEr r!rFscsg|] }tt|qSr rIrL)rNr r!rFs )r5tolistZsqueezeZsoftmaxmaxrKr3Zpost_process_object_detection) rr:r#r/rOclassesrPr- annotationZraw_annotationsZraw_annotationr )r1rNrr#r?r2r! postprocesss. "    z#ObjectDetectionPipeline.postprocessrHz torch.TensorcCs8|jdkr td|\}}}}||||d}|S)a% 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. r*z9The ObjectDetectionPipeline is only available in PyTorch.)xminyminxmaxymax)rrintrR)rrHrWrXrYrZr8r r r!r>s z)ObjectDetectionPipeline._get_bounding_box)N)r<)__name__ __module__ __qualname____doc__rr$r Predictionsr Predictionr)r7r;rVrstrr[r> __classcell__r r rr!rs   $ "-r)typingrrrrutilsrrr r r baser rZ image_utilsrr0Zmodels.auto.modeling_autorrZ get_loggerr\loggerrbrar`rr r r r!s