o oi@snddlmZmZmZmZmZddlZddlmZddl m Z ddl m Z ddl mZdgZGdddeZdS) )AnyListOptionalTupleUnionN)Tensor)PILImage)onnx) ModelBaseSuperResolutionc sPeZdZUdZdZeed<eeed<eeed<eeed<e de e e e fde e e e ffd d Z dde e e e fd ee e e e fdede e e e e dffddZ dde e e e fd ee e e e fdeededd f ddZ   d deedededee eeefdeddf fdd ZZS)!r zASuperResolution is a module that wraps an super resolution model.super_resolutionnameinput_image_sizeoutput_image_sizepseudo_image_sizeimagesreturncsX|}t|ttfr|d}n|}t|tr"fdd|D}n|}|S)aForward pass of the super resolution model. Args: images: If list of RGB images. Each image is a Tensor with shape :math:`(3, H, W)`. If Tensor, a Tensor with shape :math:`(B, 3, H, W)`. Returns: output tensor. rcsg|] }|ddqS)Nr)model).0imageselfW/home/ubuntu/.local/lib/python3.10/site-packages/kornia/models/super_resolution/base.py ?sz+SuperResolution.forward..) pre_processor isinstancelisttuplerpost_processor)rroutput out_imagesrrrforward's     zSuperResolution.forwardNtorch edge_maps output_typez Image.ImagecCs@|dur ||}g}|D]}||q |j||t|tdS)zDraw the super resolution results. Args: images: input tensor. edge_maps: detected edges. output_type: type of the output. Returns: output tensor. N)is_batch)r"append_tensor_to_typerr)rrr$r%r edge_maprrr visualizeDs   zSuperResolution.visualize directorycCs2||||}|j||dd|j||dddS)aGSave the super resolution results. Args: images: input tensor. edge_maps: detected edges. output_type: type of the output. directory: where to save outputs. output_type: backend used to generate outputs. Returns: output tensor. _src)suffix_srN)r* _save_outputs)rrr$r+r%outputsrrrsave]szSuperResolution.saveT onnx_nameinclude_pre_and_post_processorr1additional_metadatakwargszonnx.ModelProtoc s~|dur d|jd}tj|fdd|jpd|jpdgdd|jp!d|jp%dgdd|jp,d|jp0dg|r5|n|j||d|S) a?Export the current super resolution model to an ONNX model file. Args: onnx_name: The name of the output ONNX file. If not provided, a default name in the format "Kornia-.onnx" will be used. image_size: The size to which input images will be resized during preprocessing. If None, image_size will be dynamic. For DexiNed, recommended scale is 352. include_pre_and_post_processor: Whether to include the pre-processor and post-processor in the exported model. save: If to save the model or load it. additional_metadata: Additional metadata to add to the ONNX model. kwargs: Additional arguments for converting to onnx. Nkornia_z.onnxi`) input_shape output_shape pseudo_shaperr1r4)r superto_onnxrrrr)rr2r3r1r4r5 __class__rrr>us zSuperResolution.to_onnx)Nr#)NNr#)NTTN)__name__ __module__ __qualname____doc__r str__annotations__rintr#inference_moderrrr"r*r1boolrrr> __classcell__rrr?rr sd     ,  )typingrrrrrr# kornia.corerkornia.core.externalrImager kornia.models.baser __all__r rrrrs