# python test_model.py model=iphone_orig dped_dir=dped/ test_subset=full iteration=all resolution=orig use_gpu=true from scipy import misc import numpy as np import tensorflow as tf from models import resnet import utils import os import sys # process command arguments phone, dped_dir, test_subset, iteration, resolution, use_gpu = utils.process_test_model_args( sys.argv) # get all available image resolutions res_sizes = utils.get_resolutions() # get the specified image resolution IMAGE_HEIGHT, IMAGE_WIDTH, IMAGE_SIZE = utils.get_specified_res( res_sizes, phone, resolution) # disable gpu if specified config = tf.ConfigProto( device_count={'GPU': 0}) if use_gpu == "false" else None # create placeholders for input images x_ = tf.placeholder(tf.float32, [None, IMAGE_SIZE]) x_image = tf.reshape(x_, [-1, IMAGE_HEIGHT, IMAGE_WIDTH, 3]) # generate enhanced image enhanced = resnet(x_image)
import numpy as np import sys import os import time from torch.utils.data import DataLoader from torchvision import transforms import torch from load_data import LoadVisualData from model import PyNET import utils to_image = transforms.Compose([transforms.ToPILImage()]) level, restore_epoch, dataset_dir, use_gpu, orig_model = utils.process_test_model_args(sys.argv) dslr_scale = float(1) / (2 ** (level - 1)) def test_model(): if use_gpu == "true": torch.backends.cudnn.deterministic = True device = torch.device("cuda") else: os.environ['CUDA_VISIBLE_DEVICES'] = '-1' device = torch.device("cpu") # Creating dataset loaders ## visual_dataset = LoadVisualData(dataset_dir, 10, dslr_scale, level, full_resolution=True)
# python test_model.py model=iphone_orig dped_dir=dped/ test_subset=full iteration=all resolution=orig use_gpu=true import os import sys import numpy as np import tensorflow as tf from scipy import misc import models import utils # process command arguments print(sys.argv) model, dped_dir, test_subset, iteration, resolution, use_gpu, run, kernel_size, depth, blocks, parametric, s_conv, convdeconv = utils.process_test_model_args( sys.argv) dirname = model + "_" + run if run == "": dirname = model if model.endswith("_orig"): phone = model.replace("_orig", "") orig = True if phone == "iphone": dirname = "iphone_convdeconv16_iteration_40000" else: raise ValueError("No pre-trained model exists for '{}'".format(phone)) else: phone = model orig = False
# Copyright 2020 by Andrey Ignatov. All Rights Reserved. import numpy as np import tensorflow as tf import imageio import sys import os from model import PyNET import utils from load_dataset import extract_bayer_channels IMAGE_HEIGHT, IMAGE_WIDTH = 1472, 1984 LEVEL, restore_iter, dataset_dir, use_gpu, orig_model = utils.process_test_model_args( sys.argv) DSLR_SCALE = float(1) / (2**(LEVEL - 1)) # Disable gpu if specified config = tf.ConfigProto( device_count={'GPU': 0}) if use_gpu == "false" else None with tf.Session(config=config) as sess: # Placeholders for test data x_ = tf.placeholder(tf.float32, [1, IMAGE_HEIGHT, IMAGE_WIDTH, 4]) # generate enhanced image output_l0, output_l1, output_l2, output_l3, output_l4, output_l5 =\ PyNET(x_, instance_norm=True, instance_norm_level_1=False)
import numpy as np import tensorflow as tf import imageio import sys import os from model import PUNET import utils from datetime import datetime from load_dataset import extract_bayer_channels dataset_dir, test_dir, model_dir, result_dir, arch, LEVEL, inst_norm, num_maps_base,\ orig_model, rand_param, restore_iter, IMAGE_HEIGHT, IMAGE_WIDTH, use_gpu, save_model, test_image = \ utils.process_test_model_args(sys.argv) DSLR_SCALE = float(1) / (2**(max(LEVEL, 0) - 1)) # Disable gpu if specified config = tf.ConfigProto(device_count={'GPU': 0}) if not use_gpu else None with tf.compat.v1.Session(config=config) as sess: time_start = datetime.now() # determine model name if arch == "punet": name_model = "punet" # Placeholders for test data x_ = tf.compat.v1.placeholder(tf.float32, [1, IMAGE_HEIGHT // 2, IMAGE_WIDTH // 2, 4])