Clerk.Max(well);

发表于2022-08-14|图像处理

导入库1234567891011121314151617181920212223from __future__ import print_functionimport matplotlib.pyplot as plt%matplotlib inlineimport osos.environ['CUDA_VISIBLE_DEVICES'] = '1'import numpy as npfrom models.resnet import ResNetfrom models.unet import UNetfrom models.skip import skipimport torchimport torch.optimfrom utils.inpainting_utils import *torch.backends.cudnn.enabled = Truetorch.backends.cudnn.benchmark =Truedtype = torch.cuda.FloatTensorPLOT = Trueimsize = -1dim_div_by ...

Hello World

发表于2022-08-12

Deep image prior中flash_no_flash的实现

发表于2022-08-07|图像处理

导入库123456789101112131415161718192021from __future__ import print_functionimport matplotlib.pyplot as plt%matplotlib inlineimport osos.environ['CUDA_VISIBLE_DEVICES'] = '1'import numpy as npfrom models import *import torchimport torch.optimfrom utils.denoising_utils import *from utils.sr_utils import load_LR_HR_imgs_srtorch.backends.cudnn.enabled = Truetorch.backends.cudnn.benchmark =Truedtype = torch.cuda.FloatTensorimsize =-1PLOT = True 加载图片12345678imgs = load_LR_HR_imgs_sr ...

Deep image prior特征反转部分feature_inversion的实现

发表于2022-07-31|图像处理

导入库123456789101112131415161718192021222324252627from __future__ import print_functionimport matplotlib.pyplot as plt%matplotlib inlineimport argparseimport osos.environ['CUDA_VISIBLE_DEVICES'] = '0'import numpy as npfrom models import *import torchimport torch.optimfrom utils.feature_inversion_utils import *from utils.perceptual_loss.perceptual_loss import get_pretrained_netfrom utils.common_utils import *torch.backends.cudnn.enabled = Truetorch.backends.cudnn.benchmark =True ...

BSRGAN的实现

发表于2022-07-24|图像处理

测试代码123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687import os.pathimport loggingimport torchfrom utils import utils_loggerfrom utils import utils_image as utilfrom models.network_rrdbnet import RRDBNet as netdef main(): utils_logger.logger_info('blind_sr_log', log_path='blind_sr_log.log') logger = logging.getLogger('blind_sr_log') testsets = ...

Deep image prior修复部分restoration的实现

发表于2022-07-17|图像处理

导入库123456789101112131415161718192021222324252627from __future__ import print_functionimport matplotlib.pyplot as plt%matplotlib inlineimport os#os.environ['CUDA_VISIBLE_DEVICES'] = '1'import numpy as npfrom models.resnet import ResNetfrom models.unet import UNetfrom models.skip import skipfrom models import get_netimport torchimport torch.optim# from skimage.measure import compare_psnrfrom skimage.metrics import peak_signal_noise_ratio as compare_psnrfrom utils.inpainting_uti ...

Deep image prior降噪部分的实现

发表于2022-07-10|图像处理

项目源码降噪部分的修改与运行导入所需库12345678910111213141516171819202122232425from __future__ import print_functionimport matplotlib.pyplot as plt%matplotlib inlineimport os#os.environ['CUDA_VISIBLE_DEVICES'] = '3'import numpy as npfrom models import *import torchimport torch.optim# from skimage.measure import compare_psnrfrom skimage.metrics import peak_signal_noise_ratio as compare_psnrfrom utils.denoising_utils import *torch.backends.cudnn.enabled = Truetorch.backends.cudnn.benchmark =True ...

基于深度学习的图像降噪修复——文献调研

发表于2022-07-03|图像处理

基于深度学习的图像降噪修复——文献调研Deep image priorPublished in: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Date of Conference: 18-23 June 2018 Date Added to IEEE *Xplore*: 17 December 2018 ISBN Information: **Electronic ISBN:**978-1-5386-6420-9 **Print on Demand(PoD) ISBN:**978-1-5386-6421-6 ISSN Information: Electronic ISSN: 2575-7075 Print on Demand(PoD) ISSN: 1063-6919 INSPEC Accession Number: 18326119 DOI: 10.1109/CVPR.2018.00984 Publisher: IEEE Conference Location: Salt Lake City, ...