+ jH^RIt^RIHt^RIHt^RIt^RIHuHt ^RI Ht ^RI H t ^RIHt^RIHt^RIHtHt^R IHtHtHtHt^R IHt^R IHtHt^R IH t H!t!H"t"H#t#!R R]RR7t$RRlt%]!!RR] 44t&R.t'R#)N)Iterable)Union) functional)TorchvisionBackend) BatchFeature)'SemanticSegmentationPostProcessorOutput)group_images_by_shapereorder_images)ChannelDimension ImageInputPILImageResamplingSizeDict)DepthEstimatorOutput) ImagesKwargsUnpack) TensorTypeauto_docstringis_torch_availablerequires_backendsc0a]tRt^'toRtV3RltRtVtR#)CHMv2ImageProcessorKwargsa ensure_multiple_of (`int`, *optional*, defaults to 1): If `do_resize` is `True`, the image is resized to a size that is a multiple of this value. Can be overridden by `ensure_multiple_of` in `preprocess`. keep_aspect_ratio (`bool`, *optional*, defaults to `False`): If `True`, the image is resized to the largest possible size such that the aspect ratio is preserved. Can be overridden by `keep_aspect_ratio` in `preprocess`. 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. cJ<V^8dQh/S[;R&S[;R&S[;R&S[;R&#)ensure_multiple_of size_divisorkeep_aspect_ratiodo_reduce_labels)intbool)format __classdict__s"q/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/chmv2/image_processing_chmv2.py __annotate__&CHMv2ImageProcessorKwargs.__annotate__'s9 !"#N)__name__ __module__ __qualname____firstlineno____doc____annotate_func____static_attributes____classdictcell__)r!s@r"rr's r%rF)totalc xV^8dQhRRR\\\,,R\R\R\/#)r input_image torch.Tensor output_sizermultiplereturn)rrrr)r s"r"r#r#;sC$8$8$8x}$$8$8 $8  $8r%cRRlpVPRRwrVVwrxWu, p W, p V'd-\^V , 4\^V , 48dT p MT p V!W,VR7p V!W,VR7p \WR7#)rNc\W, 4V,pVe*WC8d$\P!W, 4V,pWB8d$\P!W, 4V,pV#N)roundmathfloorceil)valr4min_valmax_valxs&&&& r"constrain_to_multiple_of>get_resize_output_image_size..constrain_to_multiple_ofAsQ #. !H ,  1; 3>*X5A ; #.)H4Ar%)r4)heightwidth)rN)shapeabsr) r1r3rr4rA input_height input_width output_height output_width scale_height scale_width new_height new_widths &&&& r"get_resize_output_image_sizerP;s !, 1 1"# 6L"-M!/L,K q; #a,&6"7 7&L'K),*EPXYJ()BXVI : 77r%chaa]tRt^btoRt]t]Pt .ROt .ROt RRRR/t Rt RtRtRtRtRtRtRtRt^tRt^tV3RlV3R llt]RV3R lV3R lll4tRV3R lR lltV3RlRltV3RlRltRV3RlRlltR V3RlV3Rlllt R!V3RlRllt!RV3RlRllt"Rt#Vt$V;t%#)"CHMv2ImageProcessorz0PIL backend for CHMV2 with reduce_label support.rCirDTNFc0<V^8dQhRS[S[,/#)rkwargs)rr)r r!s"r"r# CHMv2ImageProcessor.__annotate__zs##(A!B#r%c (<\SV`!R/VBR#)Nr&)super__init__)selfrT __class__s&,r"rXCHMv2ImageProcessor.__init__zs "6"r%cP<V^8dQhRS[RS[R,RS[S[,RS[/#)rimagessegmentation_mapsNrTr5)r rrr)r r!s"r"r#rU~sD G G G&, G23 G  Gr%c &<\SV`!W3/VB#)zX segmentation_maps (`ImageInput`, *optional*): The segmentation maps to preprocess. )rW preprocess)rYr]r^rTrZs&&&,r"r`CHMv2ImageProcessor.preprocess}sw!&FvFFr%c<V^8dQhRS[RS[R,RS[RS[RS[S[,R,RS[S[R3,R,R S[/#) rr]r^Ndo_convert_rgbinput_data_formatreturn_tensorsdevicez torch.devicer5)r rr strrrr)r r!s"r"r#rUsx*C*C*C&,*C *C , *C j(4/ *Cc>)*T1*C *Cr%c VPWWFR7pVP4pRVR&/p VP!V3/VBV R&VeVPV^R\PR7p VP4p V P RRRR/4VP!R RV /V Bp V U u.uF1p V P ^4P\P4NK3 p p WR &\WR 7#uup i) z"Handle extra inputs beyond images.)r]rcrdrfFr pixel_values)r]expected_ndimsrcrd do_normalize do_rescaler]labels)data tensor_typer&) _prepare_image_like_inputscopy _preprocessr FIRSTupdatesqueezetotorchint64r) rYr]r^rcrdrerfrT images_kwargsrnprocessed_segmentation_mapssegmentation_maps_kwargsprocessed_segmentation_maps &&&&&&&, r"_preprocess_image_like_inputs1CHMv2ImageProcessor._preprocess_image_like_inputss%00L]1  ,1 ()#//H-H^  (*.*I*I( $"2"8"8 +J+ '(.{{} $ $ + +^ULRW,X Y*.*:*:+2+6N+ '3N+2M.+221588E2M (+9NBB +s%7C.cB<V^8dQhRS[R,RS[R,/#)rrmr2r5)list)r r!s"r"r#rUs%4#7D!%(%'9&7 "-"-; "5 )&<((>@TU0E^fv/w,,#% %3%9%9%; !E>!%!1!1.)!L!77 LV_N!% !M.< $U +&<**BDXYr%cH<V^8dQhRS[S[,R,RS[RR/#)r target_sizesNreturn_segmentation_scoresr5zBlist[torch.Tensor] | list[SemanticSegmentationPostProcessorOutput])rtupler)r r!s"r"r#rUs3@%@%%)%[4%7@%\`@% M@%r%c  0\4'g \R4hVPpVe\V4\V48wd \ R4h\ V\ P4'dVP4p.p\\V44Fxp\P!WF,P^R7W&,RRR7pV^,P^R7pVP\RVRV^,/R 74Kz MWVP^R7p \VP ^,4U u.uFp \RW,RWJ,/R 7NK! pp V'gVU u.uFqP"NK pp V#uup iuup i) ai Converts the output of [`CHMv2ForSemanticSegmentation`] into semantic segmentation maps. Args: outputs ([`CHMv2ForSemanticSegmentation`]): 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. return_segmentation_scores (`bool`, *optional*, defaults to `False`): Whether to return segmentation scores alongside the segmentation map. When `True`, each element of the returned list is a [`SemanticSegmentationPostProcessorOutput`] with fields `segmentation` (class IDs, shape `(height, width)`) and `segmentation_scores` (shape `(num_classes, height, width)`). Returns: `list[torch.Tensor]` or `list[SemanticSegmentationPostProcessorOutput]`: When `return_segmentation_scores=False` (default), a list of length `batch_size` where each item is a segmentation map of shape `(height, width)` with class IDs. When `return_segmentation_scores=True`, a list of [`SemanticSegmentationPostProcessorOutput`] with fields `segmentation` (class IDs, shape `(height, width)`) and `segmentation_scores` (shape `(num_classes, height, width)`). In both cases, `(height, width)` corresponds to the target size (if `target_sizes` is specified). z:PyTorch is required for post_process_semantic_segmentationzTMake sure that you pass in as many target sizes as the batch dimension of the logits)dimbilinearFrmode align_corners segmentationsegmentation_scores)rn)r ImportErrorlogitsr ValueError isinstancerwTensornumpyrF interpolate unsqueezeargmaxappendrrFr) rYoutputsrrrsemantic_segmentationrresized_logits semantic_mapseg_mapsiitems &&&& r""post_process_semantic_segmentation6CHMv2ImageProcessor.post_process_semantic_segmentations2"##Z[ [  #6{c,// j, 55+113 $& !S[)!"K))a)0|7Hzin" .a077A7> %,,;,l._get_padus9yy!45 DHH$MM%5N 0 0r%rE)rFtvFpad) rYrrrCrDrpad_top pad_bottompad_left pad_rightpaddings &&& r"rCHMv2ImageProcessor.pad_imageesP BC(  1'v<&u;i<wwu&&r%c <V^8dQhRRRS[S[S[S[S[3,,,R,R,RS[S[S[S[3,,/#)rrrrNr5)rrrrdictrg)r r!s"r"r#rUsS''''!4c3h#884?$F' d3 ?# $ 'r%c \VR4VPpVe%\V4\V48wd \R4h.pVfR.\V4,MTp\ W24F^wrVVeC\ P PPVR,VRRR7P4pVPRV/4K` V#) aj Converts the raw output of [`DepthEstimatorOutput`] into final depth predictions and depth PIL images. Only supports PyTorch. Args: outputs ([`DepthEstimatorOutput`]): Raw outputs of the model. target_sizes (`TensorType` or `List[Tuple[int, int]]`, *optional*): Tensor of shape `(batch_size, 2)` or list of tuples (`Tuple[int, int]`) containing the target size (height, width) of each image in the batch. If left to None, predictions will not be resized. Returns: `List[Dict[str, TensorType]]`: A list of dictionaries of tensors representing the processed depth predictions. rwNz]Make sure that you pass in as many target sizes as the batch dimension of the predicted depthrTrpredicted_depth)NN.) rrrrziprwnnrrrur)rYrrrresultsdepth target_sizes&&& r"post_process_depth_estimation1CHMv2ImageProcessor.post_process_depth_estimations( $(!11  $3+?3|CT+To 8D8LvO 44R^ "%o"D E&++77/*:]a8') NN-u5 6 #Er%r&)gzG?gM?gl?)g$C?g+?gM?gp?r8)NF)TF)r)&r'r(r)r*r+r valid_kwargsr BICUBICrrrrdefault_to_squarerrrrlrkrrrrrrrXrr`r}rrrrrrrr-r. __classcell__)rZr!s@@r"rRrRbs:,L!))H&J%I c7C (DIINJL FNL## G G G*C*CX9 9 v@%@%D&Z&ZP''8'''r%rR)(r:collections.abcrtypingrrwtorch.nn.functionalrrrtorchvision.transforms.v2rimage_processing_backendsrimage_processing_baserimage_processing_outputsrimage_transformsr r image_utilsr r r rmodeling_outputsrprocessing_utilsrrutilsrrrrrrPrR__all__r&r%r"rsx* $ 7;1OEUU44VV E($8NE,EEP ! !r%