update test code

README.md

@@ -0,0 +1,32 @@
+# HALDeR
+
+this is the official code for the paper "HALDER: HIERARCHICAL ATTENTION-GUIDED LEARNING WITH
+DETAIL-REFINEMENT FOR MULTI-EXPOSURE IMAGE FUSION"
+
+## Environment Preparing
+```
+python 3.6
+pytorch 1.7.0
+visdom 0.1.8.9
+dominate 2.6.0
+```
+
+### Testing
+
+We provide some example images for testing in `./test_data/`
+```
+python predict.py 
+```
+
+
+### Reference
+
+If you find our work useful in your research please consider citing our paper:
+```
+@INPROCEEDINGS{9428192,
+title={Halder: Hierarchical Attention-Guided Learning with Detail-Refinement for Multi-Exposure Image Fusion}, 
+author={Liu, Jinyuan and Shang, Jingjie and Liu, Risheng and Fan, Xin},  
+booktitle={2021 IEEE International Conference on Multimedia and Expo (ICME)}, 
+year={2021}
+}
+```

checkpoints/HALDeR/loss_log.txt

@@ -0,0 +1,13 @@
+================ Training Loss (Tue Mar 30 15:03:50 2021) ================
+================ Training Loss (Tue Mar 30 15:30:10 2021) ================
+================ Training Loss (Tue Mar 30 18:57:21 2021) ================
+================ Training Loss (Tue Mar 30 18:59:03 2021) ================
+================ Training Loss (Tue Mar 30 19:01:14 2021) ================
+================ Training Loss (Tue Mar 30 19:11:42 2021) ================
+================ Training Loss (Tue Mar 30 19:14:32 2021) ================
+================ Training Loss (Tue Mar 30 19:19:10 2021) ================
+================ Training Loss (Tue Mar 30 19:22:50 2021) ================
+================ Training Loss (Tue Mar 30 19:28:52 2021) ================
+================ Training Loss (Tue Mar 30 19:32:07 2021) ================
+================ Training Loss (Tue Mar 30 19:35:46 2021) ================
+================ Training Loss (Tue Mar 30 19:48:27 2021) ================

checkpoints/HALDeR/opt.txt

@@ -0,0 +1,83 @@
+------------ Options -------------
+D_P_times2: False
+IN_vgg: False
+aspect_ratio: 1.0
+batchSize: 1
+checkpoints_dir: ./checkpoints
+dataroot: ./test_data
+dataset_mode: test
+display_id: 1
+display_port: 8097
+display_single_pane_ncols: 0
+display_winsize: 256
+fcn: 0
+fineSize: 600
+gpu_ids: [0]
+high_times: 400
+how_many: 50
+hybrid_loss: False
+identity: 0.0
+input_linear: False
+input_nc: 3
+instance_norm: 0
+isTrain: False
+is_ca: False
+is_haze: False
+is_re: False
+is_toushe: False
+l1: 10.0
+lambda_A: 10.0
+lambda_B: 10.0
+latent_norm: False
+latent_threshold: False
+lighten: False
+linear: False
+linear_add: False
+loadSize: 286
+low_times: 200
+max_dataset_size: inf
+model: single
+multiply: False
+nThreads: 0
+n_layers_D: 3
+n_layers_patchD: 3
+name: HALDeR
+ndf: 64
+new_lr: False
+ngf: 64
+no_dropout: True
+no_flip: False
+no_vgg_instance: False
+noise: 0
+norm: instance
+norm_attention: False
+ntest: inf
+output_nc: 3
+patchD: False
+patchD_3: 0
+patchSize: 64
+patch_vgg: False
+phase: test
+resize_or_crop: no
+results_dir: ./results/
+self_attention: False
+serial_batches: False
+skip: 1
+syn_norm: False
+tanh: True
+times_residual: True
+use_avgpool: 0
+use_mse: False
+use_norm: 1
+use_ragan: False
+use_wgan: 0
+vary: 1
+vgg: 0
+vgg_choose: relu5_3
+vgg_maxpooling: False
+vgg_mean: False
+which_direction: AtoB
+which_epoch: latest
+which_model_netD: basic
+which_model_netG: sid_unet_resize
+-------------- End ----------------

data/base_data_loader.py

@@ -0,0 +1,14 @@
+
+class BaseDataLoader():
+    def __init__(self):
+        pass
+
+    def initialize(self, opt):
+        self.opt = opt
+        pass
+
+    def load_data():
+        return None
+
+
+

data/base_dataset.py

@@ -0,0 +1,48 @@
+import torch.utils.data as data
+from PIL import Image
+import torchvision.transforms as transforms
+import random
+
+class BaseDataset(data.Dataset):
+    def __init__(self):
+        super(BaseDataset, self).__init__()
+
+    def name(self):
+        return 'BaseDataset'
+
+    def initialize(self, opt):
+        pass
+
+def get_transform(opt):
+    transform_list = []
+    if opt.resize_or_crop == 'resize_and_crop':
+        zoom = 1 + 0.1*radom.randint(0,4)
+        osize = [int(400*zoom), int(600*zoom)]
+        transform_list.append(transforms.Scale(osize, Image.BICUBIC))
+        transform_list.append(transforms.RandomCrop(opt.fineSize))
+    elif opt.resize_or_crop == 'crop':
+        transform_list.append(transforms.RandomCrop(opt.fineSize))
+    elif opt.resize_or_crop == 'scale_width':
+        transform_list.append(transforms.Lambda(
+            lambda img: __scale_width(img, opt.fineSize)))
+    elif opt.resize_or_crop == 'scale_width_and_crop':
+        transform_list.append(transforms.Lambda(
+            lambda img: __scale_width(img, opt.loadSize)))
+        transform_list.append(transforms.RandomCrop(opt.fineSize))  
+    # elif opt.resize_or_crop == 'no':
+    #     osize = [384, 512]
+    #     transform_list.append(transforms.Scale(osize, Image.BICUBIC))
+
+    if opt.isTrain and not opt.no_flip:
+        transform_list.append(transforms.RandomHorizontalFlip())
+
+    transform_list += [transforms.ToTensor()]
+    return transforms.Compose(transform_list)
+
+def __scale_width(img, target_width):
+    ow, oh = img.size
+    if (ow == target_width):
+        return img
+    w = target_width
+    h = int(target_width * oh / ow)
+    return img.resize((w, h), Image.BICUBIC)

data/custom_dataset_data_loader.py

@@ -0,0 +1,50 @@
+import torch.utils.data
+from data.base_data_loader import BaseDataLoader
+
+
+def CreateDataset(opt):
+    dataset = None
+    if opt.dataset_mode == 'aligned':
+        from data.aligned_dataset import AlignedDataset
+        dataset = AlignedDataset()
+    elif opt.dataset_mode == 'unaligned':
+        from data.unaligned_dataset import UnalignedDataset
+        dataset = UnalignedDataset()
+    elif opt.dataset_mode == 'unaligned_random_crop':
+        from data.unaligned_random_crop import UnalignedDataset
+        dataset = UnalignedDataset()
+    elif opt.dataset_mode == 'pair':
+        from data.pair_dataset import PairDataset
+        dataset = PairDataset()
+    elif opt.dataset_mode == 'test':
+        from data.test_dataset import PairDataset
+        dataset = PairDataset()
+    elif opt.dataset_mode == 'single':
+        from data.single_dataset import SingleDataset
+        dataset = SingleDataset()
+    else:
+        raise ValueError("Dataset [%s] not recognized." % opt.dataset_mode)
+
+    print("dataset [%s] was created" % (dataset.name()))
+    dataset.initialize(opt)
+    return dataset
+
+
+class CustomDatasetDataLoader(BaseDataLoader):
+    def name(self):
+        return 'CustomDatasetDataLoader'
+
+    def initialize(self, opt):
+        BaseDataLoader.initialize(self, opt)
+        self.dataset = CreateDataset(opt)
+        self.dataloader = torch.utils.data.DataLoader(
+            self.dataset,
+            batch_size=opt.batchSize,
+            shuffle=not opt.serial_batches,
+            num_workers=int(opt.nThreads))
+
+    def load_data(self):
+        return self.dataloader
+
+    def __len__(self):
+        return min(len(self.dataset), self.opt.max_dataset_size)

data/data_loader.py

@@ -0,0 +1,7 @@
+
+def CreateDataLoader(opt):
+    from data.custom_dataset_data_loader import CustomDatasetDataLoader
+    data_loader = CustomDatasetDataLoader()
+    print(data_loader.name())
+    data_loader.initialize(opt)
+    return data_loader