+ jAxRt^RIt^RIHt^RIHtHt^RIt^RIHt^RI H t !RR]P4t R#)z, Implements the Generalized R-CNN framework N) OrderedDict)OptionalUnion)nn)_log_api_usage_oncecaa]tRt^toRtV3RlV3Rllt]PPV3RlRl4t R V3RlRllt Rt Vt V;t #) GeneralizedRCNNa@ Main class for Generalized R-CNN. Args: backbone (nn.Module): rpn (nn.Module): roi_heads (nn.Module): takes the features + the proposals from the RPN and computes detections / masks from it. transform (nn.Module): performs the data transformation from the inputs to feed into the model c <V^8dQhRS[PRS[PRS[PRS[PRR/#)backbonerpn roi_heads transformreturnN)rModule)format __classdict__s"n/home/luca/screen-checking/.venv/lib/python3.14/site-packages/torchvision/models/detection/generalized_rcnn.py __annotate__GeneralizedRCNN.__annotate__sM!!))!YY!99 ! 99 !  !cx<\SV`4\V4W@nWnW nW0nRVnR#)FN)super__init__rrr r r _has_warned)selfr r r r __class__s&&&&&rrGeneralizedRCNN.__init__s4 D!" " rc  <V^8dQhRS[S[S[P3,RS[S[S[S[P3,,RS[S[S[S[P3,S[S[S[S[P3,,3,/#)r losses detectionsr)dictstrtorchTensorlistr)rrs"rrr-sq3 ,-;?S%,,EV@W;X tC%&T#u||2C-D(EE Frc.VP'dV#V#N)training)rrr s&&&r eager_outputsGeneralizedRCNN.eager_outputs,s ===Mrc <V^8dQhRS[S[P,RS[S[S[S[S[P3,,,RS[S[S[S[P3,S[S[S[S[P3,,3,/#)r imagestargetsr)r%r#r$rr!r"tuple)rrs"rrr5szP:P:U\\"P:$tC$5678P: tC%&T#u||2C-D(EE F P:rc  VP'dVf\P!RR4MVFpVR,p\V\P4'd\\P!\ VP 4^8H;'dVP R,^8HRVP R24K\P!RR\V4 R24K .pVFfpVP RRp\P!\ V4^8HRVP RR 24VPV^,V^,34Kh VPW4wrVe\V4FwrVR,pVR,VR,8*p V P4'gK9\P!V P^R 74^,^,p WJ,P4p \P!RR V R V R24K VPVP4p \V \P4'd\!R V 3.4p VP#WV4wrVP%WVP&V4wppVPP)WP&V4p/pVP+V4VP+V4\P,P/4'd4VP0'g\2P4!R 4RVnVV3#VP7VV4#)a Args: images (list[Tensor]): images to be processed targets (list[dict[str, tensor]]): ground-truth boxes present in the image (optional) Returns: result (list[BoxList] or dict[Tensor]): the output from the model. During training, it returns a dict[Tensor] which contains the losses. During testing, it returns list[BoxList] contains additional fields like `scores`, `labels` and `mask` (for Mask R-CNN models). NFz0targets should not be none when in training modeboxesz:Expected target boxes to be a tensor of shape [N, 4], got .z0Expected target boxes to be of type Tensor, got zJexpecting the last two dimensions of the Tensor to be H and W instead got )dimzLAll bounding boxes should have positive height and width. Found invalid box z for target at index 0z=RCNN always returns a (Losses, Detections) tuple in scriptingT)NNN:r NN)r6:Nr N)r(r#_assert isinstancer$lenshapetypeappendr enumerateanywheretolistr tensorsrr r image_sizes postprocessupdatejit is_scriptingrwarningswarnr))rr,r-targetr0original_image_sizesimgval target_idxdegenerate_boxesbb_idxdegen_bbfeatures proposalsproposal_lossesr detector_lossesrs&&& rforwardGeneralizedRCNN.forward5s" === e%WX%F"7OE!%66  ,1JJekk"o6JXY^YdYdXeefg  !NtTY{m[\]&79C))BC.C MMCA \]`]f]fgigj]k\lm  ! ' 'QQ(8 9 ..9  &/&8" w#(<5<#? #''))"[[)9)=)=!)=)DEaHKF,1M,@,@,BHMM..6Z7LZLXY['9==0 h - -"S(O#45H%)XXf%H" &*nnX&J\J\^e&f# O^^// **,@  o& o& 99 ! ! # #### ]^#' :% %%%fj9 9r)rr r r rr')__name__ __module__ __qualname____firstlineno____doc__rr#rEunusedr)rU__static_attributes____classdictcell__ __classcell__)rrs@@rrrsD !!  YYP:P:P:rr) r[rG collectionsrtypingrrr#rutilsrrrrrrds0#" (v:biiv:r