a  þd¤ã@sJddlZddlZddlmZdd„Zddd„Zdd „Zd d „Zd d „Z dS)éN)Ú functionalcCs4tj |¡}|dkr|S|dkr(| ¡S| ¡SdS)zËReduce loss as specified. Args: loss (Tensor): Elementwise loss tensor. reduction (str): Options are 'none', 'mean' and 'sum'. Returns: Tensor: Reduced loss tensor. réN)ÚFZ _ReductionZget_enumÚmeanÚsum)ÚlossÚ reductionZreduction_enum©r úa/var/www/html/stable-diffusion-webui/venv/lib/python3.9/site-packages/basicsr/losses/loss_util.pyÚ reduce_losss r rcCs¨|durJ| ¡| ¡ksJ‚| d¡dksB| d¡| d¡ksBJ‚||}|dusZ|dkrft||ƒ}n>|dkr¤| d¡dkr†| ¡}n| ¡| d¡}| ¡|}|S)aMApply element-wise weight and reduce loss. Args: loss (Tensor): Element-wise loss. weight (Tensor): Element-wise weights. Default: None. reduction (str): Same as built-in losses of PyTorch. Options are 'none', 'mean' and 'sum'. Default: 'mean'. Returns: Tensor: Loss values. Nrrr)ÚdimÚsizer r)rÚweightrr r r Úweight_reduce_losss &   rcst ˆ¡d‡fdd„ ƒ}|S)aÄCreate a weighted version of a given loss function. To use this decorator, the loss function must have the signature like `loss_func(pred, target, **kwargs)`. The function only needs to compute element-wise loss without any reduction. This decorator will add weight and reduction arguments to the function. The decorated function will have the signature like `loss_func(pred, target, weight=None, reduction='mean', **kwargs)`. :Example: >>> import torch >>> @weighted_loss >>> def l1_loss(pred, target): >>> return (pred - target).abs() >>> pred = torch.Tensor([0, 2, 3]) >>> target = torch.Tensor([1, 1, 1]) >>> weight = torch.Tensor([1, 0, 1]) >>> l1_loss(pred, target) tensor(1.3333) >>> l1_loss(pred, target, weight) tensor(1.5000) >>> l1_loss(pred, target, reduction='none') tensor([1., 1., 2.]) >>> l1_loss(pred, target, weight, reduction='sum') tensor(3.) Nrcs"ˆ||fi|¤Ž}t|||ƒ}|S)N)r)ÚpredÚtargetrrÚkwargsr©Ú loss_funcr r ÚwrapperYs zweighted_loss..wrapper)Nr)Ú functoolsÚwraps)rrr rr Ú weighted_loss:srcCs^|dd}tj|||||gdd}| d|d¡ d|d¡}tj|dddd d ¡ d ¡}|S) agGet local weights for generating the artifact map of LDL. It is only called by the `get_refined_artifact_map` function. Args: residual (Tensor): Residual between predicted and ground truth images. ksize (Int): size of the local window. Returns: Tensor: weight for each pixel to be discriminated as an artifact pixel réZreflect)ÚpadÚmodeé)éÿÿÿÿéþÿÿÿT)r ZunbiasedÚkeepdimr)rrZunfoldÚtorchÚvarZsqueeze)ZresidualÚksizerZ residual_padZunfolded_residualÚpixel_level_weightr r r Úget_local_weightscs r$c Csrtjt ||¡ddd}tjt ||¡ddd}tj| ¡dddd}t| ¡|ƒ}||}d|||k<|S)a8Calculate the artifact map of LDL (Details or Artifacts: A Locally Discriminative Learning Approach to Realistic Image Super-Resolution. In CVPR 2022) Args: img_gt (Tensor): ground truth images. img_output (Tensor): output images given by the optimizing model. img_ema (Tensor): output images given by the ema model. ksize (Int): size of the local window. Returns: overall_weight: weight for each pixel to be discriminated as an artifact pixel (calculated based on both local and global observations). rT)r)rréýÿÿÿ)r rgš™™™™™É?r)r rÚabsr!Úcloner$) Zimg_gtZ img_outputZimg_emar"Z residual_emaZ residual_srZpatch_level_weightr#Zoverall_weightr r r Úget_refined_artifact_mapys r()Nr) rr Ztorch.nnrrr rrr$r(r r r r Ús  )