Yh2,dddlZddlmZddlZddlZddlmZmZddlmZddl m Z ddl m Z ddl mZmZdd lmZd d lmZmZd Zd ZdZeddZejj ddedeeeefde ededededefdZGddej Z!dS) N)Union)nnTensor)is_compile_supported)BroadcastingList2)_pair)_assert_has_ops_has_ops)_log_api_usage_once)check_roi_boxes_shapeconvert_boxes_to_roi_formatc fd}|S)zkLazily wrap a function with torch.compile on the first call This avoids eagerly importing dynamo. cJtjfd}|S)Nctjfi}tj|t j<||i|SN)torchcompile functoolswrapsglobals__name__)argskwargs compiled_fncompile_kwargsfns k/var/www/tools.fuzzalab.pt/emblema-extractor/venv/lib/python3.11/site-packages/torchvision/ops/roi_align.py compile_hookz7lazy_compile..decorate_fn..compile_hooksS-==n==K%8Y_R%8%8%E%EGIIbk ";/// /)rr)rr rs` r decorate_fnz!lazy_compile..decorate_fns?    0 0 0 0 0   0 r!)rr"s` r lazy_compiler$s$  r!c\}}}|d}|d}|} |} tj| |dz k|dz | dz} tj| |dz k|dz | } tj| |dz k|j|}tj| |dz k|dz | dz} tj| |dz k|dz | } tj| |dz k|j|}|| z } || z }d| z }d|z }fd}|| | }|| | }|| | }|| | }d}|||}|||}|| |}|| |}||z||zz||zz||zz}|S)Nrminr ?c fJJtjdddddf|d}tjdddddf|d}dddddddftjjdddddddf|dddddddddf|dddddddddffS)Nrdevice)rwherearanger+)yxchannelsinput roi_batch_indxmaskymasks r masked_indexz+_bilinear_interpolate..masked_indexBs  $$$ E!!!T111*-q!44A E!!!T111*-q!44A !!!T4tT9 : L%, 7 7 7aaatTSW8W X aaaqqq$4' ( aaatQQQaaa' ( *  r!c T|dddddddddf|dddddddddfzSrr#)r.r/s r outer_prodz)_bilinear_interpolate..outer_prodWsID!!!T111d*+a4qqq$0I.JJJr!)sizeclampintrr,todtype)r1r2r.r/r4r3_heightwidthy_lowx_lowy_highx_highlylxhyhxr5v1v2v3v4r7w1w2w3w4valr0s`` `` @r_bilinear_interpolaterQ#sQ"'Ax AA AA EEGGE EEGGE [&1*,fqj%!) D DF K!+VaZ ? ?E EVaZ'ek):):A>>A [%!)+UQY B BF K*EAIu = =E EUQY&U[(9(91==A UB UB rB rB           eU # #B eV $ $B fe $ $B ff % %BKKK B  B B  B B  B B  B r'BG b2g %R /C Jr!ctjr0|jr)|jtjkr|S|Sr)ris_autocast_enabledis_cudar<doublefloat)tensors r maybe_castrXes=  ""v~&,%,:V:V||~~ r!T)dynamicc ^|j}t|}t|}|\}}} } tj||j} tj||j} |dddf} |rdnd}|dddf|z|z }|dddf|z|z }|dddf|z|z }|dddf|z|z }||z }||z }|s,tj|d }tj|d }||z }||z }|dk}|r|ntj||z }|r|ntj||z } |rNt||zd}tj||j}tj||j}d}d}ntj||zd }tj| |j}tj| |j}|dddf|dddfk}|dddf|dddfk}d }||| ddddf||zz|ddddfdz |j|||z zz} ||| ddddf||zz|ddddfdz |j|||z zz}!t|| | |!||}"|sLtj |ddddddddf|"d}"tj |ddddddddf|"d}"|" d }#t|tjr|#|dddddfz}#n|#|z}#|# |}#|#S) Nr*rg?gr r r(r&c|ddddfSrr#)ts rfrom_Kz_roi_align..from_KsD$r!))r<rXr8rr-r+r:r9ceilmaxr;rQr,sum isinstancer)$r1rois spatial_scale pooled_height pooled_widthsampling_ratioaligned orig_dtyper=r>r?phpwr2offset roi_start_w roi_start_h roi_end_w roi_end_h roi_width roi_height bin_size_h bin_size_wexact_samplingroi_bin_grid_hroi_bin_grid_wcountiyixr4r3r_r.r/rPoutputs$ r _roi_alignrrsgJ u  E d  D**,,Aq&% mEL 9 9 9B l5< 8 8 8B AJNN$$M $SSFqqq!t*},v5Kqqq!t*},v5KQQQT ]*V3IQQQT ]*V3IK'I[(J 6K s333 [555 m+J\)J#a'N'5a^^5:jS`F`;a;aN'5_^^5:iR^F^;_;_N6N^3Q77 \. > > > \. > > > N^;CCC \& 6 6 6 \% 5 5 547 nQQQW5547 nQQQW55    { T111d] ffZ00 0 1 dD!!!m s " & &u{ 3 3ffZ.=X6Y6Y Y Z { T111d] ffZ00 0 1 dD!!!m s " & &u{ 3 3ffZ.=X6Y6Y Y Z }aE5 I IC Gk%4tQQQ <=sAFFk%4tT111 <=sAFF WWX  F%&&%4t+,,% YYz " "F Mr!r(r`Fr1boxes output_sizergrjrkreturnc tjs2tjst t t ||}t|}t|tj st|}tjsptr(tj rO|j s|js|jr:t!|jjr!t'||||d|d||St)tjj||||d|d||S)aj Performs Region of Interest (RoI) Align operator with average pooling, as described in Mask R-CNN. Args: input (Tensor[N, C, H, W]): The input tensor, i.e. a batch with ``N`` elements. Each element contains ``C`` feature maps of dimensions ``H x W``. If the tensor is quantized, we expect a batch size of ``N == 1``. boxes (Tensor[K, 5] or List[Tensor[L, 4]]): the box coordinates in (x1, y1, x2, y2) format where the regions will be taken from. The coordinate must satisfy ``0 <= x1 < x2`` and ``0 <= y1 < y2``. If a single Tensor is passed, then the first column should contain the index of the corresponding element in the batch, i.e. a number in ``[0, N - 1]``. If a list of Tensors is passed, then each Tensor will correspond to the boxes for an element i in the batch. output_size (int or Tuple[int, int]): the size of the output (in bins or pixels) after the pooling is performed, as (height, width). spatial_scale (float): a scaling factor that maps the box coordinates to the input coordinates. For example, if your boxes are defined on the scale of a 224x224 image and your input is a 112x112 feature map (resulting from a 0.5x scaling of the original image), you'll want to set this to 0.5. Default: 1.0 sampling_ratio (int): number of sampling points in the interpolation grid used to compute the output value of each pooled output bin. If > 0, then exactly ``sampling_ratio x sampling_ratio`` sampling points per bin are used. If <= 0, then an adaptive number of grid points are used (computed as ``ceil(roi_width / output_width)``, and likewise for height). Default: -1 aligned (bool): If False, use the legacy implementation. If True, pixel shift the box coordinates it by -0.5 for a better alignment with the two neighboring pixel indices. This version is used in Detectron2 Returns: Tensor[K, C, output_size[0], output_size[1]]: The pooled RoIs. rr )rjit is_scripting is_tracingr roi_alignrrrerrr $are_deterministic_algorithms_enabledrTis_mpsis_xpurr+typerr ops torchvision)r1rrrgrjrkrfs rrrsXR 9 ! ! # #'EI,@,@,B,B'I&&&%   D $$K dEL ) )1*400 9 ! ! # #s  s:<< sBG- sSXS_ schco s#5<#455 seT=+a.+VW.Zhjqrr r 9 * * t]KNKNNT[  r!c eZdZdZ d deedededeffd Zde d e e e e fd e fd Z d e fd ZxZS)RoIAlignz See :func:`roi_align`. Frrgrjrkctt|||_||_||_||_dSr)super__init__r rrgrjrk)selfrrgrjrk __class__s rrzRoIAlign.__init__ sL D!!!&*, r!r1rfrcRt|||j|j|j|jSr)rrrgrjrk)rr1rfs rforwardzRoIAlign.forwards'd&68JDL_aeamnnnr!c d|jjd|jd|jd|jd|jd }|S)Nz (output_size=z, spatial_scale=z, sampling_ratio=z , aligned=))rrrrgrjrk)rss r__repr__zRoIAlign.__repr__sh~&  +  #1  !% 3       r!)F)r __module__ __qualname____doc__rr:rVboolrrrlistrstrr __classcell__)rs@rrrs  &s+          oVo5f1E+Fo6oooo #        r!r)r(r`F)"rtypingrrtorch.fxrrtorch._dynamo.utilsrtorch.jit.annotationsrtorch.nn.modules.utilsrtorchvision.extensionr r utilsr _utilsrrr$rQrXrfxwraprr:rVrrModulerr#r!rrs 444444333333((((((;;;;;;;;''''''FFFFFFFF&===DdUUUp  88 8 f% &8#3'8 8  8  8 8888vryr!