`i\DdZddlmZmZddlZddlmZmZm Z ddl m Z m Z m Z mZddlmZmZmZmZmZmZmZmZmZddlmZmZmZmZmZmZdd l m!Z!erddl"Z"erddl#Z#ej$e%Z&e!d Gd d eZ'd gZ(dS)z$Image processor class for MobileViT.)OptionalUnionN)BaseImageProcessor BatchFeature get_size_dict)flip_channel_orderget_resize_output_image_sizeresizeto_channel_dimension_format) ChannelDimension ImageInputPILImageResamplinginfer_channel_dimension_formatis_scaled_imagemake_flat_list_of_imagesto_numpy_array valid_imagesvalidate_preprocess_arguments) TensorTypefilter_out_non_signature_kwargsis_torch_availableis_torch_tensoris_vision_availablelogging)requires)vision)backendsc!eZdZdZdgZddejddddddf dedee e e fd ed ed e e e fd ed ee e e fdededdffd Zejddfdejde e e fd edee e efdee e efdejf dZ d$dejdee e efdee e efdejfdZdedejfdZd%fd Z d&dededed ed ededee e e fd eed ee d ee e e fdee e effdZ d'dedeedee e e fd eed eed ee d eed ee e e fdeedee e efdee e efdejfdZ d(dedeedeedee e e fd eed ee e e fdee e efdejfdZedddddddddddejdf dedeedeedee e e fd eed eed ee d eed ee e e fdeedeed ee e efdedee e efdejjfd!Zd%d"ee e!fd#Z"xZ#S))MobileViTImageProcessora Constructs a MobileViT image processor. Args: do_resize (`bool`, *optional*, defaults to `True`): Whether to resize the image's (height, width) dimensions to the specified `size`. Can be overridden by the `do_resize` parameter in the `preprocess` method. size (`dict[str, int]` *optional*, defaults to `{"shortest_edge": 224}`): Controls the size of the output image after resizing. Can be overridden by the `size` parameter in the `preprocess` method. resample (`PILImageResampling`, *optional*, defaults to `Resampling.BILINEAR`): Defines the resampling filter to use if resizing the image. Can be overridden by the `resample` parameter in the `preprocess` method. do_rescale (`bool`, *optional*, defaults to `True`): Whether to rescale the image by the specified scale `rescale_factor`. Can be overridden by the `do_rescale` parameter in the `preprocess` method. rescale_factor (`int` or `float`, *optional*, defaults to `1/255`): Scale factor to use if rescaling the image. Can be overridden by the `rescale_factor` parameter in the `preprocess` method. do_center_crop (`bool`, *optional*, defaults to `True`): Whether to crop the input at the center. If the input size is smaller than `crop_size` along any edge, the image is padded with 0's and then center cropped. Can be overridden by the `do_center_crop` parameter in the `preprocess` method. crop_size (`dict[str, int]`, *optional*, defaults to `{"height": 256, "width": 256}`): Desired output size `(size["height"], size["width"])` when applying center-cropping. Can be overridden by the `crop_size` parameter in the `preprocess` method. do_flip_channel_order (`bool`, *optional*, defaults to `True`): Whether to flip the color channels from RGB to BGR. Can be overridden by the `do_flip_channel_order` parameter in the `preprocess` method. do_reduce_labels (`bool`, *optional*, defaults to `False`): Whether or not to reduce all label values of segmentation maps by 1. Usually used for datasets where 0 is used for background, and background itself is not included in all classes of a dataset (e.g. ADE20k). The background label will be replaced by 255. Can be overridden by the `do_reduce_labels` parameter in the `preprocess` method. pixel_valuesTNgp?F do_resizesizeresample do_rescalerescale_factordo_center_crop crop_sizedo_flip_channel_orderdo_reduce_labelsreturnc  tjd i| ||nddi}t|d}||nddd}t|d}||_||_||_||_||_||_||_ ||_ | |_ dS) N shortest_edgeFdefault_to_square)heightwidthr( param_name) super__init__rr"r#r$r%r&r'r(r)r*) selfr"r#r$r%r&r'r(r)r*kwargs __class__s /home/jaya/work/projects/VOICE-AGENT/VIET/agent-env/lib/python3.11/site-packages/transformers/models/mobilevit/image_processing_mobilevit.pyr8z MobileViTImageProcessor.__init___s ""6"""'ttos-CTU;;;!*!6IIsUXr#r$r?r@r:r0 output_sizes r<r zMobileViTImageProcessor.resize}s2! d " "(D %     'T//NDM2DD`aa a2 //      #/       r=c&t|||S)a Flip the color channels from RGB to BGR or vice versa. Args: image (`np.ndarray`): The image, represented as a numpy array. data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format of the image. If not provided, it will be the same as the input image. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format of the input image. If not provided, it will be inferred. )r?r@)r )r9r>r?r@s r<r z*MobileViTImageProcessor.flip_channel_orders""%[Teffffr=labelcRt|}d||dk<|dz }d||dk<|S)Nr)r)r9rEs r< reduce_labelz$MobileViTImageProcessor.reduce_labels9u%%eqj !esl r=c <tj|fd|i|S)z Preprocesses a batch of images and optionally segmentation maps. Overrides the `__call__` method of the `Preprocessor` class so that both images and segmentation maps can be passed in as positional arguments. segmentation_maps)r7__call__)r9imagesrLr:r;s r<rMz MobileViTImageProcessor.__call__s* uwwVV:KVvVVVr=c |r||}|r||||| }|r||| | }|r||| | }|r||| }|S)N)r>r#r$r@)r>scaler@)r>r#r@)r@)rJr rescale center_cropr ) r9r>r*r"r%r'r)r#r$r&r(r@s r< _preprocessz#MobileViTImageProcessor._preprocesss  -%%e,,E  pKKe$]nKooE  iLLuNVgLhhE  g$$5yTe$ffE  X++EEV+WWE r=c t|}|r)t|rtd| t |} ||d|||||||| |  }t || | }|S)zPreprocesses a single image.zIt looks like you are trying to rescale already rescaled images. If the input images have pixel values between 0 and 1, set `do_rescale=False` to avoid rescaling them again.NF) r>r*r"r#r$r%r&r'r(r)r@)input_channel_dim)rrlogger warning_oncerrSr ) r9r>r"r#r$r%r&r'r(r)r?r@s r<_preprocess_imagez)MobileViTImageProcessor._preprocess_images u%%  /%00    s     $ >u E E   "!))"7/!   ,E;Rcddd r=segmentation_mapc Pt|}|jdkrd}|d}tj}nd}|t |d}|||||t jd||d| }|r|d }| tj }|S) zPreprocesses a single mask.T)N.FNrH) num_channels) r>r*r"r#r$r%r'r(r)r@r) rndimr FIRSTrrSrNEARESTsqueezeastypenpint64) r9rYr*r"r#r'r(r@added_channel_dims r<_preprocess_maskz(MobileViTImageProcessor._preprocess_masks**:;;  A % % $ / :  0 6   %  ($BCSbc$d$d$d!++"-'/)"'/,     ;/77:: +2228<<r=rNrLreturn_tensorsc  njnjnjnjnj  nj njtd  nj t d  nj t|}|t|d}t|}t|std|t|std t    f d |D}d |i}|  fd |D}||d<t|| S)a Preprocess an image or batch of images. Args: images (`ImageInput`): Image to preprocess. Expects a single or batch of images with pixel values ranging from 0 to 255. If passing in images with pixel values between 0 and 1, set `do_rescale=False`. segmentation_maps (`ImageInput`, *optional*): Segmentation map to preprocess. do_resize (`bool`, *optional*, defaults to `self.do_resize`): Whether to resize the image. size (`dict[str, int]`, *optional*, defaults to `self.size`): Size of the image after resizing. resample (`int`, *optional*, defaults to `self.resample`): Resampling filter to use if resizing the image. This can be one of the enum `PILImageResampling`, Only has an effect if `do_resize` is set to `True`. do_rescale (`bool`, *optional*, defaults to `self.do_rescale`): Whether to rescale the image by rescale factor. rescale_factor (`float`, *optional*, defaults to `self.rescale_factor`): Rescale factor to rescale the image by if `do_rescale` is set to `True`. do_center_crop (`bool`, *optional*, defaults to `self.do_center_crop`): Whether to center crop the image. crop_size (`dict[str, int]`, *optional*, defaults to `self.crop_size`): Size of the center crop if `do_center_crop` is set to `True`. do_flip_channel_order (`bool`, *optional*, defaults to `self.do_flip_channel_order`): Whether to flip the channel order of the image. do_reduce_labels (`bool`, *optional*, defaults to `self.do_reduce_labels`): Whether or not to reduce all label values of segmentation maps by 1. Usually used for datasets where 0 is used for background, and background itself is not included in all classes of a dataset (e.g. ADE20k). The background label will be replaced by 255. return_tensors (`str` or `TensorType`, *optional*): The type of tensors to return. Can be one of: - Unset: Return a list of `np.ndarray`. - `TensorType.TENSORFLOW` or `'tf'`: Return a batch of type `tf.Tensor`. - `TensorType.PYTORCH` or `'pt'`: Return a batch of type `torch.Tensor`. - `TensorType.NUMPY` or `'np'`: Return a batch of type `np.ndarray`. - `TensorType.JAX` or `'jax'`: Return a batch of type `jax.numpy.ndarray`. data_format (`ChannelDimension` or `str`, *optional*, defaults to `ChannelDimension.FIRST`): The channel dimension format for the output image. Can be one of: - `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `ChannelDimension.LAST`: image in (height, width, num_channels) format. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format for the input image. If unset, the channel dimension format is inferred from the input image. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - `"none"` or `ChannelDimension.NONE`: image in (height, width) format. NFr/r(r4r[)expected_ndimszkInvalid image type. Must be of type PIL.Image.Image, numpy.ndarray, torch.Tensor, tf.Tensor or jax.ndarray.zvInvalid segmentation map type. Must be of type PIL.Image.Image, numpy.ndarray, torch.Tensor, tf.Tensor or jax.ndarray.)r%r&r'r(r"r#r$cP g|]"} |   #S)) r>r"r#r$r%r&r'r(r)r?r@)rX) .0imgr(r?r'r)r%r"r@r$r&r9r#s r< z6MobileViTImageProcessor.preprocess..sa     " "#!%--#&;'"3 #     r=r!c Hg|]}|S))rYr*r"r#r'r(r@)re) rjrYr(r'r*r"r@r9r#s r<rlz6MobileViTImageProcessor.preprocess..sU ! ! !%%%%5%5'#1'&7& ! ! !r=labels)data tensor_type)r"r$r%r&r'r)r#rr(r*rrrBrr)r9rNrLr"r#r$r%r&r'r(r)r*rfr?r@ros` ````````` `` r< preprocessz"MobileViTImageProcessor.preprocessEs`D"+!6IIDN '388#-#9ZZt +9+E4K^+9+E4K^%:%F ! !DLf 'ttTYTU;;;!*!6IIDN !) DDD /?/K++QUQf)&11  ( 89J[\ ] ] ] )&11F## :   (>O1P1P (:  &!))                     "'  ( ! ! ! ! ! ! ! ! ! !): ! ! ! /DN>BBBBr= target_sizescl|j}|t|t|krtdt|r|}gt t|D]{}t jj || d||dd}|d d} ||n<| dfd t j dDS) a@ Converts the output of [`MobileViTForSemanticSegmentation`] into semantic segmentation maps. Only supports PyTorch. Args: outputs ([`MobileViTForSemanticSegmentation`]): Raw outputs of the model. target_sizes (`list[Tuple]` of length `batch_size`, *optional*): List of tuples corresponding to the requested final size (height, width) of each prediction. If unset, predictions will not be resized. Returns: semantic_segmentation: `list[torch.Tensor]` of length `batch_size`, where each item is a semantic segmentation map of shape (height, width) corresponding to the target_sizes entry (if `target_sizes` is specified). Each entry of each `torch.Tensor` correspond to a semantic class id. NzTMake sure that you pass in as many target sizes as the batch dimension of the logitsr)dimbilinearF)r#mode align_cornersrHc g|] }| Sr6r6)rjisemantic_segmentations r<rlzNMobileViTImageProcessor.post_process_semantic_segmentation..s$m$m$m!%:1%=$m$m$mr=)logitslenrBrnumpyrangetorchnn functional interpolate unsqueezeargmaxappendshape)r9outputsrrr{idxresized_logits semantic_maprzs @r<"post_process_semantic_segmentationz:MobileViTImageProcessor.post_process_semantic_segmentationsM"  #6{{c,//// j|,, 4+1133 $& !S[[)) ; ;!&!4!@!@3K))a)00|C7Hzin"A"" .a077A7>> %,,\::::  ;%+MMaM$8$8 !$m$m$m$muMbMhijMkGlGl$m$m$m !$$r=)NN)N)NNNNN) NNNNNNNNNN)NNNNNN)$__name__ __module__ __qualname____doc__model_input_namesrBILINEARboolrdictstrintrfloatr8rbndarrayr r r rrJrMrSrXrerr^rPILImagerqlisttupler __classcell__)r;s@r<r r 7sq""H(()-'9'B,3#.2&*!&111tCH~&1% 1  1 c5j) 11DcN+1 $11 111111D(:'B>BDH / / z/ 38n/ % / eC)9$9:; / $E#/?*?$@A /  / / / / h?CDH ggzgeC)9$9:;g$E#/?*?$@A g  gggg(*WWWWWW"*.15*..2DH      $tCH~&-.!DcN+$E#/?*?$@AD%))-15%)*.)-.204>BDH)))D>)tCH~& ) -. ) TN )!)!)DcN+) (~)eC)9$9:;)$E#/?*?$@A) ))))\,0$()-)-.2DH& & $& #4.& D> & tCH~& & ! & DcN+& $E#/?*?$@A&  & & & & P%$&&37$()-15%)*.)-.204+/;?(8(>DHQCQCQC$J/QCD> QC tCH~& QC -. QCTNQC!QC!QCDcN+QC (~QC#4.QC!sJ!78QC&QC$E#/?*?$@AQC !QCQCQC'&QCh)%)%QUV[Q\H])%)%)%)%)%)%)%)%r=r ))rtypingrrr}rbimage_processing_utilsrrrimage_transformsr r r r image_utilsr rrrrrrrrutilsrrrrrrutils.import_utilsrrr get_loggerrrVr __all__r6r=r<rs+*""""""""UUUUUUUUUUuuuuuuuuuuuu                      +*****JJJLLL  H % % ;K%K%K%K%K%0K%K% K%\ % %r=