3jSrSSKJr SSKJr SSKrSSKJs Jr SSK J r J r SSK Jr S SjrS S SjjrSSS jjrSSS jjrg)zImplementation of "differentiable spatial to numerical" (soft-argmax) operations. As described in the paper "Numerical Coordinate Regression with Convolutional Neural Networks" by Nibali et al. ) annotations)OptionalNKORNIA_CHECK_IS_TENSORKORNIA_CHECK_SHAPE)create_meshgridc6[U5 [U/SQ5 g)N)BCHWr)tensors U/home/wildlama/miniconda3/lib/python3.13/site-packages/kornia/geometry/subpix/dsnt.py$_validate_batched_image_tensor_inputr"s6"v34c&[U5 URup#pEUc[R"S5nUR UR UR S9nURX#S5n[R"Xa-SS9nURX#XE5$)aApply the Softmax function over features in each image channel. Note that this function behaves differently to :py:class:`torch.nn.Softmax2d`, which instead applies Softmax over features at each spatial location. Args: input: the input torch.Tensor with shape :math:`(B, N, H, W)`. temperature: factor to apply to input, adjusting the "smoothness" of the output distribution. Returns: a 2D probability distribution per image channel with shape :math:`(B, N, H, W)`. Examples: >>> heatmaps = torch.tensor([[[ ... [0., 0., 0.], ... [0., 0., 0.], ... [0., 1., 2.]]]]) >>> spatial_softmax2d(heatmaps) tensor([[[[0.0585, 0.0585, 0.0585], [0.0585, 0.0585, 0.0585], [0.0585, 0.1589, 0.4319]]]]) g?devicedtype)dim) rshapetorchrtorrviewFsoftmax)input temperature batch_sizechannelsheightwidthxx_softs rspatial_softmax2dr&'s{0)/*/++'J&ll3' .. EKK.HK :,A YYqB /F ;;zV ;;rc[U5 URup#pE[XEXR5nUR UR 5nUSR S5nUSR S5nURX#S5n [R"X-SSS9n [R"Xy-SSS9n [R"X/S5n U RX#S5$)aVCompute the expectation of coordinate values using spatial probabilities. The input heatmap is assumed to represent a valid spatial probability distribution, which can be achieved using :func:`~kornia.geometry.subpixel.spatial_softmax2d`. Args: input: the input torch.Tensor representing dense spatial probabilities with shape :math:`(B, N, H, W)`. normalized_coordinates: whether to return the coordinates normalized in the range of :math:`[-1, 1]`. Otherwise, it will return the coordinates in the range of the input shape. Returns: expected value of the 2D coordinates with shape :math:`(B, N, 2)`. Output order of the coordinates is (x, y). Examples: >>> heatmaps = torch.tensor([[[ ... [0., 0., 0.], ... [0., 0., 0.], ... [0., 1., 0.]]]]) >>> spatial_expectation2d(heatmaps, False) tensor([[[1., 2.]]]) .rr.T)keepdim) rrrrrrreshaperrsumcat) rnormalized_coordinatesr r!r"r#gridpos_xpos_y input_flat expected_y expected_xoutputs rspatial_expectation2dr8Ls.)/*/++'J& 6*@,, OD 775;; D L  $E L  $EJ"5J5-r4@J5-r4@J YY / 4F ;;zQ //rcURUR:XaURUR:Xd [S5eUupEURnURnU(a/[R"SSXWUS9n[R"SSXGUS9n O4[R"SUS- XWUS9n[R"SUS- XGUS9n USR S5n USR S5n USR S5n USR S5n X- S-nX- S-nS [R "U S-5-nS [R "U S-5-n[R"UU-5n[R"UU-5nUURSS S 9S -- nUURSS S 9S -- nUR S5UR S 5-$)aRender the PDF of a 2D Gaussian distribution. Args: mean: the mean location of the Gaussian to render, :math:`(\mu_x, \mu_y)`. Shape: :math:`(*, 2)`. std: the standard deviation of the Gaussian to render, :math:`(\sigma_x, \sigma_y)`. Shape :math:`(*, 2)`. Should be able to be broadcast with `mean`. size: the (height, width) of the output image. normalized_coordinates: whether ``mean`` and ``std`` are assumed to use coordinates normalized in the range of :math:`[-1, 1]`. Otherwise, coordinates are assumed to be in the range of the output shape. Returns: torch.Tensor including rendered points with shape :math:`(*, H, W)`. z1Expected inputs to have the same dtype and devicerr*rrr(r)r,gT)rr+g:0yE>) rr TypeErrorrlinspace unsqueeze reciprocalexpr.)meanstdsizer0r"r#rrxsysmu_xmu_ysigma_xsigma_y dist_x_sq dist_y_sqk_xk_ygauss_xgauss_ys rrender_gaussian2drOys" II # dkk(AKLLMF JJE [[F ^^B5u E ^^B6 F ^^Auqy%e L ^^Avz6 N < ! !" %D < ! !" %D&k##B'G&k##B'Gq Iq I !!'1*- -C !!'1*- -Cii C(Gii C(GT:TABGT:TABG   R 7#4#4R#8 88r)r torch.TensorreturnNone)N)rrPrzOptional[torch.Tensor]rQrP)T)rrPr0boolrQrP) r@rPrArPrBztuple[int, int]r0rSrQrP)__doc__ __future__rtypingrrtorch.nn.functionalnn functionalrkornia.core.checkrrkornia.geometry.gridrrr&r8rOrrr]sf$ # H05 "