添加image光流

.gitignore

@@ -1,5 +0,0 @@
-checkpoints/
-*.log
-*.pth
-*.ckpt
-__pycache__/

checkpoints/ROMA_UNSB_001/loss_log.txt

@@ -0,0 +1,80 @@
+================ Training Loss (Sun Feb 23 15:46:44 2025) ================
+================ Training Loss (Sun Feb 23 15:52:29 2025) ================
+================ Training Loss (Sun Feb 23 16:00:07 2025) ================
+================ Training Loss (Sun Feb 23 16:02:40 2025) ================
+================ Training Loss (Sun Feb 23 16:05:19 2025) ================
+================ Training Loss (Sun Feb 23 16:06:44 2025) ================
+================ Training Loss (Sun Feb 23 16:09:38 2025) ================
+================ Training Loss (Sun Feb 23 16:44:56 2025) ================
+================ Training Loss (Sun Feb 23 16:49:46 2025) ================
+================ Training Loss (Sun Feb 23 16:51:03 2025) ================
+================ Training Loss (Sun Feb 23 16:51:23 2025) ================
+================ Training Loss (Sun Feb 23 18:04:02 2025) ================
+================ Training Loss (Sun Feb 23 18:04:39 2025) ================
+================ Training Loss (Sun Feb 23 18:05:17 2025) ================
+================ Training Loss (Sun Feb 23 18:06:40 2025) ================
+================ Training Loss (Sun Feb 23 18:11:48 2025) ================
+================ Training Loss (Sun Feb 23 18:13:31 2025) ================
+================ Training Loss (Sun Feb 23 18:14:11 2025) ================
+================ Training Loss (Sun Feb 23 18:14:29 2025) ================
+================ Training Loss (Sun Feb 23 18:16:27 2025) ================
+================ Training Loss (Sun Feb 23 18:16:44 2025) ================
+================ Training Loss (Sun Feb 23 18:20:39 2025) ================
+================ Training Loss (Sun Feb 23 18:21:44 2025) ================
+================ Training Loss (Sun Feb 23 18:35:27 2025) ================
+================ Training Loss (Sun Feb 23 18:39:21 2025) ================
+================ Training Loss (Sun Feb 23 18:40:15 2025) ================
+================ Training Loss (Sun Feb 23 18:41:15 2025) ================
+================ Training Loss (Sun Feb 23 18:47:46 2025) ================
+================ Training Loss (Sun Feb 23 18:48:36 2025) ================
+================ Training Loss (Sun Feb 23 18:50:20 2025) ================
+================ Training Loss (Sun Feb 23 18:51:50 2025) ================
+================ Training Loss (Sun Feb 23 18:58:45 2025) ================
+================ Training Loss (Sun Feb 23 18:59:52 2025) ================
+================ Training Loss (Sun Feb 23 19:03:05 2025) ================
+================ Training Loss (Sun Feb 23 19:03:57 2025) ================
+================ Training Loss (Sun Feb 23 21:11:47 2025) ================
+================ Training Loss (Sun Feb 23 21:17:10 2025) ================
+================ Training Loss (Sun Feb 23 21:20:14 2025) ================
+================ Training Loss (Sun Feb 23 21:29:03 2025) ================
+================ Training Loss (Sun Feb 23 21:34:57 2025) ================
+================ Training Loss (Sun Feb 23 21:35:26 2025) ================
+================ Training Loss (Sun Feb 23 22:28:43 2025) ================
+================ Training Loss (Sun Feb 23 22:29:04 2025) ================
+================ Training Loss (Sun Feb 23 22:29:52 2025) ================
+================ Training Loss (Sun Feb 23 22:30:40 2025) ================
+================ Training Loss (Sun Feb 23 22:33:48 2025) ================
+================ Training Loss (Sun Feb 23 22:39:16 2025) ================
+================ Training Loss (Sun Feb 23 22:39:48 2025) ================
+================ Training Loss (Sun Feb 23 22:41:34 2025) ================
+================ Training Loss (Sun Feb 23 22:42:01 2025) ================
+================ Training Loss (Sun Feb 23 22:44:17 2025) ================
+================ Training Loss (Sun Feb 23 22:45:53 2025) ================
+================ Training Loss (Sun Feb 23 22:46:48 2025) ================
+================ Training Loss (Sun Feb 23 22:47:42 2025) ================
+================ Training Loss (Sun Feb 23 22:49:44 2025) ================
+================ Training Loss (Sun Feb 23 22:50:29 2025) ================
+================ Training Loss (Sun Feb 23 22:51:47 2025) ================
+================ Training Loss (Sun Feb 23 22:55:56 2025) ================
+================ Training Loss (Sun Feb 23 22:56:19 2025) ================
+================ Training Loss (Sun Feb 23 22:57:58 2025) ================
+================ Training Loss (Sun Feb 23 22:59:09 2025) ================
+================ Training Loss (Sun Feb 23 23:02:36 2025) ================
+================ Training Loss (Sun Feb 23 23:03:56 2025) ================
+================ Training Loss (Sun Feb 23 23:09:21 2025) ================
+================ Training Loss (Sun Feb 23 23:10:05 2025) ================
+================ Training Loss (Sun Feb 23 23:11:43 2025) ================
+================ Training Loss (Sun Feb 23 23:12:41 2025) ================
+================ Training Loss (Sun Feb 23 23:13:05 2025) ================
+================ Training Loss (Sun Feb 23 23:13:59 2025) ================
+================ Training Loss (Sun Feb 23 23:14:59 2025) ================
+================ Training Loss (Mon Feb 24 21:53:50 2025) ================
+================ Training Loss (Mon Feb 24 21:54:16 2025) ================
+================ Training Loss (Mon Feb 24 21:54:50 2025) ================
+================ Training Loss (Mon Feb 24 21:55:31 2025) ================
+================ Training Loss (Mon Feb 24 21:56:10 2025) ================
+================ Training Loss (Mon Feb 24 22:09:38 2025) ================
+================ Training Loss (Mon Feb 24 22:10:16 2025) ================
+================ Training Loss (Mon Feb 24 22:12:46 2025) ================
+================ Training Loss (Mon Feb 24 22:13:04 2025) ================
+================ Training Loss (Mon Feb 24 22:14:04 2025) ================

checkpoints/ROMA_UNSB_001/train_opt.txt

@@ -0,0 +1,88 @@
+----------------- Options ---------------
+            adj_size_list: [2, 4, 6, 8, 12]              
+             atten_layers: 1,3,5                         
+               batch_size: 1                             
+                    beta1: 0.5                           
+                    beta2: 0.999                         
+          checkpoints_dir: ./checkpoints                 
+           continue_train: False                         
+                crop_size: 256                           
+                 dataroot: /home/openxs/kunyu/datasets/InfraredCity-Lite/Double/Moitor	[default: placeholder]
+             dataset_mode: unaligned_double              	[default: unaligned]
+                direction: AtoB                          
+              display_env: ROMA                          	[default: main]
+             display_freq: 50                            
+               display_id: None                          
+            display_ncols: 4                             
+             display_port: 8097                          
+           display_server: http://localhost              
+          display_winsize: 256                           
+               easy_label: experiment_name               
+                    epoch: latest                        
+              epoch_count: 1                             
+                eta_ratio: 0.1                           
+          evaluation_freq: 5000                          
+        flip_equivariance: False                         
+                 gan_mode: lsgan                         
+                  gpu_ids: 0                             
+                init_gain: 0.02                          
+                init_type: xavier                        
+                 input_nc: 3                             
+                  isTrain: True                          	[default: None]
+             lambda_D_ViT: 1.0                           
+               lambda_GAN: 8.0                           	[default: 1.0]
+               lambda_NCE: 8.0                           	[default: 1.0]
+                lambda_SB: 0.1                           
+               lambda_ctn: 1.0                           
+            lambda_global: 1.0                           
+               lambda_inc: 1.0                           
+                   lmda_1: 0.1                           
+                load_size: 286                           
+                       lr: 1e-05                         	[default: 0.0002]
+           lr_decay_iters: 50                            
+                lr_policy: linear                        
+         max_dataset_size: inf                           
+                    model: roma_unsb                     	[default: cut]
+                 n_epochs: 100                           
+           n_epochs_decay: 100                           
+               n_layers_D: 3                             
+                    n_mlp: 3                             
+                     name: ROMA_UNSB_001                 	[default: experiment_name]
+                    nce_T: 0.07                          
+                  nce_idt: False                         	[default: True]
+nce_includes_all_negatives_from_minibatch: False                         
+               nce_layers: 0,4,8,12,16                   
+                      ndf: 64                            
+                     netD: basic_cond                    
+                     netF: mlp_sample                    
+                  netF_nc: 256                           
+                     netG: resnet_9blocks_cond           
+                      ngf: 64                            
+             no_antialias: False                         
+          no_antialias_up: False                         
+               no_dropout: True                          
+                  no_flip: True                          	[default: False]
+                  no_html: False                         
+                    normD: instance                      
+                    normG: instance                      
+              num_patches: 256                           
+              num_threads: 4                             
+            num_timesteps: 10                            	[default: 5]
+                output_nc: 3                             
+                    phase: train                         
+                pool_size: 0                             
+               preprocess: resize_and_crop               
+          pretrained_name: None                          
+               print_freq: 100                           
+         random_scale_max: 3.0                           
+             save_by_iter: False                         
+          save_epoch_freq: 5                             
+         save_latest_freq: 5000                          
+           serial_batches: False                         
+stylegan2_G_num_downsampling: 1                             
+                   suffix:                               
+                      tau: 0.01                          
+         update_html_freq: 1000                          
+                  use_idt: False                         
+                  verbose: False                         
+----------------- End -------------------

models/roma_unsb_model.py

@@ -60,15 +60,11 @@ def compute_ctn_loss(G, x, F_content): #公式10
     loss = F.mse_loss(warped_fake, y_fake_warped)
     return loss
 
-class ContentAwareOptimization(nn.Module):
+class ContentAwareOptimization(nn.Module): 
     def __init__(self, lambda_inc=2.0, eta_ratio=0.4):
         super().__init__()
         self.lambda_inc = lambda_inc  # 权重增强系数
         self.eta_ratio = eta_ratio    # 选择内容区域的比例
-        
-        # 改为类成员变量，确保钩子函数可访问
-        self.gradients_real = []
-        self.gradients_fake = []
     
     def compute_cosine_similarity(self, gradients):
         """
@@ -82,66 +78,138 @@ class ContentAwareOptimization(nn.Module):
         # 计算余弦相似度
         cosine_sim = F.cosine_similarity(gradients, mean_grad, dim=2)  # [B, N]
         return cosine_sim
-
-    def generate_weight_map(self, gradients_real, gradients_fake):
+    
+    def generate_weight_map(self, gradients_fake, feature_shape):
         """
-        生成内容感知权重图
+        生成内容感知权重图（修正空间维度）
         Args:
             gradients_real: [B, N, D] 真实图像判别器梯度
             gradients_fake: [B, N, D] 生成图像判别器梯度
+            feature_shape: tuple [H, W] 判别器输出的特征图尺寸
         Returns:
-            weight_real: [B, N] 真实图像权重图
-            weight_fake: [B, N] 生成图像权重图
+            weight_real: [B, 1, H, W] 真实图像权重图
+            weight_fake: [B, 1, H, W] 生成图像权重图
         """
-        # 计算真实图像块的余弦相似度
-        cosine_real = self.compute_cosine_similarity(gradients_real)  # [B, N]  公式5
-        # 计算生成图像块的余弦相似度
-        cosine_fake = self.compute_cosine_similarity(gradients_fake)  # [B, N]
+        H, W = feature_shape
+        N = H * W
         
-        # 选择内容丰富的区域（余弦相似度最低的eta_ratio比例）
-        k = int(self.eta_ratio * cosine_real.shape[1])
+        # 计算余弦相似度（与原代码相同）
+        cosine_fake = self.compute_cosine_similarity(gradients_fake)
         
-        # 对真实图像生成权重图
-        _, real_indices = torch.topk(-cosine_real, k, dim=1)  # 选择最不相似的区域
-        weight_real = torch.ones_like(cosine_real)
-        for b in range(cosine_real.shape[0]):
-            weight_real[b, real_indices[b]] = self.lambda_inc / (1e-6 + torch.abs(cosine_real[b, real_indices[b]])) #公式6
-        
-        # 对生成图像生成权重图（同理）
+        # 生成权重图（与原代码相同）
+        k = int(self.eta_ratio * cosine_fake.shape[1])
         _, fake_indices = torch.topk(-cosine_fake, k, dim=1)
         weight_fake = torch.ones_like(cosine_fake)
+    
         for b in range(cosine_fake.shape[0]):
-            weight_fake[b, fake_indices[b]] = self.lambda_inc / (1e-6 + torch.abs(cosine_fake[b, fake_indices[b]]))
+            weight_fake[b, fake_indices[b]] = self.lambda_inc / (1e-6 + torch.abs(cosine_fake[b, fake_indices[b]]))    
+    
+        # 重建空间维度 --------------------------------------------------
+        # 将权重从[B, N]转换为[B, H, W]
+        #print(f"Shape of weight_fake before view: {weight_fake.shape}")
+        #print(f"Shape of cosine_fake: {cosine_fake.shape}")
+        #print(f"H: {H}, W: {W}, N: {N}")
+        weight_fake = weight_fake.view(-1, H, W).unsqueeze(1) # [B,1,H,W]
         
-        return weight_real, weight_fake
+        return  weight_fake
+
+    def compute_cosine_similarity_image(self, gradients):
+        """
+        计算每个空间位置梯度与平均梯度的余弦相似度 (图像版本)
+        Args:
+            gradients: [B, C, H, W] 判别器输出的梯度
+        Returns:
+            cosine_sim: [B, H, W] 每个空间位置的余弦相似度
+        """
+        # 将空间维度展平，以便计算所有空间位置的平均梯度
+        B, C, H, W = gradients.shape
+        gradients_reshaped = gradients.view(B, C, H * W) # [B, C, N] where N = H*W
+        gradients_transposed = gradients_reshaped.transpose(1, 2) # [B, N, C]  将C放到最后一维，方便计算空间位置的平均梯度
+
+        mean_grad = torch.mean(gradients_transposed, dim=1, keepdim=True)  # [B, 1, C]  在空间位置维度上求平均，得到平均梯度 [B, 1, C]
+        # mean_grad 现在是所有空间位置的平均梯度，形状为 [B, 1, C]
+
+        # 为了计算余弦相似度，我们需要将 mean_grad 扩展到与 gradients_transposed 相同的空间维度
+        mean_grad_expanded = mean_grad.expand(-1, H * W, -1) # [B, N, C]
+
+        # 计算余弦相似度，dim=2 表示在特征维度 (C) 上计算
+        cosine_sim = F.cosine_similarity(gradients_transposed, mean_grad_expanded, dim=2)  # [B, N]
+
+        # 将 cosine_sim 重新reshape回 [B, H, W]
+        cosine_sim = cosine_sim.view(B, H, W)
+        return cosine_sim
+
+    def generate_weight_map_image(self, gradients_fake, feature_shape):
+        """
+        生成内容感知权重图（修正空间维度 - 图像版本）
+        Args:
+            gradients_fake: [B, C, H, W] 生成图像判别器梯度
+            feature_shape: tuple [H, W] 判别器输出的特征图尺寸
+        Returns:
+            weight_fake: [B, 1, H, W] 生成图像权重图
+        """
+        H, W = feature_shape
+        # 计算余弦相似度（图像版本）
+        cosine_fake = self.compute_cosine_similarity_image(gradients_fake) # [B, H, W]
+        # 生成权重图（与原代码相同，但现在cosine_fake是[B, H, W]）
+        k = int(self.eta_ratio * H * W) # k 仍然是基于总的空间位置数量计算
+        _, fake_indices = torch.topk(-cosine_fake.view(cosine_fake.shape[0], -1), k, dim=1) # 将 cosine_fake 展平为 [B, N] 以使用 topk
+        weight_fake = torch.ones_like(cosine_fake).view(cosine_fake.shape[0], -1) # 初始化权重图，并展平为 [B, N]
+        for b in range(cosine_fake.shape[0]):
+            weight_fake[b, fake_indices[b]] = self.lambda_inc / (1e-6 + torch.abs(cosine_fake.view(cosine_fake.shape[0], -1)[b, fake_indices[b]]))
+        weight_fake = weight_fake.view(-1, H, W).unsqueeze(1) # 重新 reshape 为 [B, H, W]，并添加通道维度变为 [B, 1, H, W]
+        return  weight_fake
 
     def forward(self, D_real, D_fake, real_scores, fake_scores):
-        # 清空梯度缓存
-        self.gradients_real.clear()
-        self.gradients_fake.clear()
+        """
+        计算内容感知对抗损失
+        Args:
+            D_real: 判别器对真实图像的特征输出 [B, C, H, W]
+            D_fake: 判别器对生成图像的特征输出 [B, C, H, W]
+            real_scores: 真实图像的判别器预测 [B, N] (N=H*W)
+            fake_scores: 生成图像的判别器预测 [B, N]
+        Returns:
+            loss_co_adv: 内容感知对抗损失
+        """
+        B, C, H, W = D_real.shape
+        N = H * W
+        shape_hw = [H, W]
+        # 注册钩子获取梯度
+        gradients_real = []
+        gradients_fake = []
         
-        # 注册钩子
-        hook_real = lambda grad: self.gradients_real.append(grad.detach())
-        hook_fake = lambda grad: self.gradients_fake.append(grad.detach())
+        def hook_real(grad):
+            gradients_real.append(grad.detach().view(B, N, -1))
+        
+        def hook_fake(grad):
+            gradients_fake.append(grad.detach().view(B, N, -1))
         
         D_real.register_hook(hook_real)
         D_fake.register_hook(hook_fake)
         
-        # 触发梯度计算
-        (real_scores.mean() + fake_scores.mean()).backward(retain_graph=True)
+        # 计算原始对抗损失以触发梯度计算
+        loss_real = torch.mean(torch.log(real_scores + 1e-8))
+        loss_fake = torch.mean(torch.log(1 - fake_scores + 1e-8))
+        # 添加与 D_real、D_fake 相关的 dummy 项，确保梯度传递
+        loss_dummy = 1e-8 * (D_real.sum() + D_fake.sum())
+        total_loss = loss_real + loss_fake + loss_dummy
+        total_loss.backward(retain_graph=True)
         
-        # 获取梯度并调整维度
-        grad_real = self.gradients_real[0]  # [B, N, D]
-        grad_fake = self.gradients_fake[0]
+        # 获取梯度数据
+        gradients_real = gradients_real[1]  # [B, N, D]
+        gradients_fake = gradients_fake[1]  # [B, N, D]
         
         # 生成权重图
-        weight_real, weight_fake = self.generate_weight_map(grad_real, grad_fake)
+        self.weight_real, self.weight_fake = self.generate_weight_map(gradients_fake, shape_hw )
         
-        # 计算加权损失
-        loss_co_real = (weight_real * torch.log(real_scores + 1e-8)).mean()
-        loss_co_fake = (weight_fake * torch.log(1 - fake_scores + 1e-8)).mean()
+        # 应用权重到对抗损失
+        loss_co_real = torch.mean(self.weight_real * torch.log(real_scores + 1e-8))
+        loss_co_fake = torch.mean(self.weight_fake * torch.log(1 - fake_scores + 1e-8))
         
-        return -(loss_co_real + loss_co_fake), weight_real, weight_fake
+        # 计算并返回最终内容感知对抗损失
+        loss_co_adv = -(loss_co_real + loss_co_fake)
+        
+        return loss_co_adv
 
 class ContentAwareTemporalNorm(nn.Module):
     def __init__(self, gamma_stride=0.1, kernel_size=21, sigma=5.0):
@@ -149,33 +217,6 @@ class ContentAwareTemporalNorm(nn.Module):
         self.gamma_stride = gamma_stride  # 控制整体运动幅度
         self.smoother = GaussianBlur(kernel_size, sigma=sigma)  # 高斯平滑层
 
-    def upsample_weight_map(self, weight_patch, target_size=(256, 256)):
-        """
-        将patch级别的权重图上采样到目标分辨率
-        Args:
-            weight_patch: [B, 1, 24, 24] 来自ViT的patch权重图
-            target_size: 目标分辨率 (H, W)
-        Returns:
-            weight_full: [B, 1, 256, 256] 上采样后的全分辨率权重图
-        """
-        # 使用双线性插值上采样
-        B = weight_patch.shape[0]
-        weight_patch = weight_patch.view(B, 1, 24, 24)
-        
-        weight_full = F.interpolate(
-            weight_patch,
-            size=target_size,
-            mode='bilinear',
-            align_corners=False
-        )
-        
-        # 对每个16x16的patch内部保持权重一致（可选）
-        # 通过平均池化再扩展，消除插值引入的渐变
-        weight_full = F.avg_pool2d(weight_full, kernel_size=16, stride=16)
-        weight_full = F.interpolate(weight_full, scale_factor=16, mode='nearest')
-        
-        return weight_full
-    
     def forward(self, weight_map):
         """
         生成内容感知光流
@@ -184,16 +225,15 @@ class ContentAwareTemporalNorm(nn.Module):
         Returns:
             F_content: [B, 2, H, W] 生成的光流场(x/y方向位移)
         """
-        # 上采样权重图到全分辨率
-        weight_full = self.upsample_weight_map(weight_map)  # [B,1,384,384]     
+        print(weight_map.shape)
+        B, _, H, W = weight_map.shape
         
         # 1. 归一化权重图
         # 保持区域相对强度，同时限制数值范围
-        weight_norm = F.normalize(weight_full, p=1, dim=(2,3))  # L1归一化 [B,1,H,W]
+        weight_norm = F.normalize(weight_map, p=1, dim=(2,3))  # L1归一化 [B,1,H,W]
         
-        # 2. 生成高斯噪声
-        B, _, H, W = weight_norm.shape
-        z = torch.randn(B, 2, H, W, device=weight_norm.device)  # [B,2,H,W]
+        # 2. 生成高斯噪声(与光流场同尺寸)
+        z = torch.randn(B, 2, H, W, device=weight_map.device)  # [B,2,H,W]
         
         # 3. 合成基础光流
         # 将权重图扩展为2通道(x/y方向共享权重)
@@ -223,25 +263,29 @@ class RomaUnsbModel(BaseModel):
         parser.add_argument('--lambda_global', type=float, default=1.0, help='weight for Global Structural Consistency')
         
         parser.add_argument('--nce_idt', type=util.str2bool, nargs='?', const=True, default=False, help='use NCE loss for identity mapping: NCE(G(Y), Y))')
+        parser.add_argument('--nce_layers', type=str, default='0,4,8,12,16', help='compute NCE loss on which layers')
         parser.add_argument('--nce_includes_all_negatives_from_minibatch',
                             type=util.str2bool, nargs='?', const=True, default=False,
                             help='(used for single image translation) If True, include the negatives from the other samples of the minibatch when computing the contrastive loss. Please see models/patchnce.py for more details.')
-        parser.add_argument('--nce_layers', type=str, default='0,4,8,12,16', help='compute NCE loss on which layers')
         
         parser.add_argument('--netF', type=str, default='mlp_sample', choices=['sample', 'reshape', 'mlp_sample'], help='how to downsample the feature map')
+        parser.add_argument('--netF_nc', type=int, default=256)
+        parser.add_argument('--nce_T', type=float, default=0.07, help='temperature for NCE loss')
         
+        parser.add_argument('--lmda_1', type=float, default=0.1)
+        parser.add_argument('--num_patches', type=int, default=256, help='number of patches per layer')
         parser.add_argument('--flip_equivariance',
                             type=util.str2bool, nargs='?', const=True, default=False,
                             help="Enforce flip-equivariance as additional regularization. It's used by FastCUT, but not CUT")
         
         parser.add_argument('--lambda_inc', type=float, default=1.0, help='incremental weight for content-aware optimization')
-        parser.add_argument('--eta_ratio', type=float, default=0.4, help='ratio of content-rich regions')
+        parser.add_argument('--eta_ratio', type=float, default=0.1, help='ratio of content-rich regions')
 
         parser.add_argument('--atten_layers', type=str, default='5', help='compute Cross-Similarity on which layers')
         
         parser.add_argument('--tau', type=float, default=0.01, help='Entropy parameter')
         parser.add_argument('--num_timesteps', type=int, default=5, help='# of discrim filters in the first conv layer')
-        
+        parser.add_argument('--adj_size_list', type=list, default=[2, 4, 6, 8, 12], help='different scales of perception field')
         parser.add_argument('--n_mlp', type=int, default=3, help='only used if netD==n_layers')
 
         parser.set_defaults(pool_size=0)  # no image pooling
@@ -258,8 +302,9 @@ class RomaUnsbModel(BaseModel):
         BaseModel.__init__(self, opt)
         
         # 指定需要打印的训练损失
-        self.loss_names = ['G_GAN', 'D_real_ViT', 'D_fake_ViT', 'G', 'SB']
-        self.visual_names = ['real_A0', 'real_A_noisy', 'fake_B0', 'real_B0']
+        self.loss_names = ['G_GAN_1', 'D_real_1', 'D_fake_1', 'G_1', 'NCE_1', 'SB_1', 
+                           'G_2']
+        self.visual_names = ['real_A', 'real_A_noisy', 'fake_B', 'real_B']
         self.atten_layers = [int(i) for i in self.opt.atten_layers.split(',')]
         
         if self.opt.phase == 'test':
@@ -275,10 +320,12 @@ class RomaUnsbModel(BaseModel):
 
         if self.isTrain:
             self.model_names = ['G', 'D_ViT', 'E']
+
             
         else:
             self.model_names = ['G']
         
+        print(f'input_nc = {self.opt.input_nc}')
         # 创建网络
         self.netG = networks.define_G(opt.input_nc, opt.output_nc, opt.ngf, opt.netG, opt.normG, not opt.no_dropout, opt.init_type, opt.init_gain, opt.no_antialias, opt.no_antialias_up, self.gpu_ids, opt)
         
@@ -296,6 +343,9 @@ class RomaUnsbModel(BaseModel):
             # 定义损失函数
             self.criterionL1 = torch.nn.L1Loss().to(self.device)
             self.criterionGAN = networks.GANLoss(opt.gan_mode).to(self.device)
+            self.criterionNCE = []
+            for nce_layer in self.nce_layers:
+                self.criterionNCE.append(PatchNCELoss(opt).to(self.device))
             self.criterionIdt = torch.nn.L1Loss().to(self.device)
             self.optimizer_G = torch.optim.Adam(self.netG.parameters(), lr=opt.lr, betas=(opt.beta1, opt.beta2))
             self.optimizer_D = torch.optim.Adam(self.netD_ViT.parameters(), lr=opt.lr, betas=(opt.beta1, opt.beta2))
@@ -373,7 +423,6 @@ class RomaUnsbModel(BaseModel):
         self.real_B1 = input['B1' if AtoB else 'A1'].to(self.device)
         self.image_paths = input['A_paths' if AtoB else 'B_paths']
         
-        
     def tokens_concat(self, origin_tokens, adjacent_size):
         adj_size = adjacent_size
         B, token_num, C = origin_tokens.shape[0], origin_tokens.shape[1], origin_tokens.shape[2]
@@ -394,10 +443,9 @@ class RomaUnsbModel(BaseModel):
                 cut_patch = torch.mean(cut_patch, dim=1, keepdim=True)
                 cut_patch_list.append(cut_patch)
 
-
         result = torch.cat(cut_patch_list,dim=1)
         return result
-        
+    
     def cat_results(self, origin_tokens, adj_size_list):
         res_list = [origin_tokens]
         for ad_s in adj_size_list:
@@ -405,7 +453,6 @@ class RomaUnsbModel(BaseModel):
             res_list.append(cat_result)
         
         result = torch.cat(res_list, dim=1)
-
         return result
     
     def forward(self):
@@ -456,8 +503,8 @@ class RomaUnsbModel(BaseModel):
 
         # ============ 第三步：拼接输入并执行网络推理 =============
         bs = self.real_A0.size(0)
-        self.z_in = torch.randn(size=[bs, 4 * self.opt.ngf]).to(self.real_A0.device)
-        self.z_in2 = torch.randn(size=[bs, 4 * self.opt.ngf]).to(self.real_A1.device)
+        z_in = torch.randn(size=[bs, 4 * self.opt.ngf]).to(self.real_A0.device)
+        z_in2 = torch.randn(size=[bs, 4 * self.opt.ngf]).to(self.real_A1.device)
         # 将 real_A, real_B 拼接 (如 nce_idt=True)，并同样处理 real_A_noisy 与 XtB
         self.real = self.real_A0
         self.realt = self.real_A_noisy
@@ -468,8 +515,8 @@ class RomaUnsbModel(BaseModel):
                 self.real = torch.flip(self.real, [3])
                 self.realt = torch.flip(self.realt, [3])
         
-        self.fake_B0 = self.netG(self.real_A0, self.time, self.z_in)
-        self.fake_B1 = self.netG(self.real_A1, self.time, self.z_in2)
+        self.fake_B0 = self.netG(self.real_A0, self.time, z_in)
+        self.fake_B1 = self.netG(self.real_A1, self.time, z_in2)
 
         if self.opt.phase == 'train':
             real_A0 = self.real_A0
@@ -494,7 +541,32 @@ class RomaUnsbModel(BaseModel):
             self.mutil_fake_B1_tokens = self.netPreViT(self.fake_B1_resize, self.atten_layers, get_tokens=True)
             # [[1,576,768],[1,576,768],[1,576,768]]
             #  [3,576,768]
+            #self.mutil_real_A0_tokens = self.cat_results(self.mutil_real_A0_tokens[0], self.opt.adj_size_list)
+            #print(f'self.mutil_real_A0_tokens[0]:{self.mutil_real_A0_tokens[0].shape}')
             
+            shape_hw = list(self.real_A0_resize.shape[2:4])
+            # 生成图像的梯度
+            fake_gradient = torch.autograd.grad(self.mutil_fake_B0_tokens[0].sum(), self.mutil_fake_B0_tokens, create_graph=True)[0]
+            
+            # 梯度图
+            self.weight_fake = self.cao.generate_weight_map_image(fake_gradient, shape_hw)
+            
+            # 生成图像的CTN光流图
+            self.f_content = self.ctn(self.weight_fake)
+            
+            # 变换后的图片
+            self.warped_real_A_noisy2 = warp(self.real_A_noisy, self.f_content)
+            self.warped_fake_B0       = warp(self.fake_B0,self.f_content)
+            
+            # 经过第二次生成器
+            self.warped_fake_B0_2 = self.netG(self.warped_real_A_noisy2, self.time, z_in)   
+            
+            # warped_fake_B0_2=self.warped_fake_B0_2
+            # warped_fake_B0=self.warped_fake_B0
+            # self.warped_fake_B0_2_resize = self.resize(warped_fake_B0_2)
+            # self.warped_fake_B0_resize = self.resize(warped_fake_B0)
+            # self.mutil_warped_fake_B0_tokens = self.netPreViT(self.warped_fake_B0_resize, self.atten_layers, get_tokens=True)
+            # self.mutil_fake_B0_2_tokens = self.netPreViT(self.warped_fake_B0_2_resize, self.atten_layers, get_tokens=True)
 
             
     def compute_D_loss(self):      #判别器还是没有改
@@ -502,23 +574,30 @@ class RomaUnsbModel(BaseModel):
 
         lambda_D_ViT = self.opt.lambda_D_ViT
         fake_B0_tokens = self.mutil_fake_B0_tokens[0].detach()
-       
+        fake_B1_tokens = self.mutil_fake_B1_tokens[0].detach()
 
         real_B0_tokens = self.mutil_real_B0_tokens[0]
+        real_B1_tokens = self.mutil_real_B1_tokens[0]
+
 
         pre_fake0_ViT = self.netD_ViT(fake_B0_tokens)
-        self.loss_D_fake_ViT = self.criterionGAN(pre_fake0_ViT, False)
+        pre_fake1_ViT = self.netD_ViT(fake_B1_tokens)
 
-        pred_real0_ViT = self.netD_ViT(real_B0_tokens)       
-        self.loss_D_real_ViT = self.criterionGAN(pred_real0_ViT, True)
+        self.loss_D_fake_ViT = (self.criterionGAN(pre_fake0_ViT, False).mean() + self.criterionGAN(pre_fake1_ViT, False).mean()) * 0.5 * lambda_D_ViT
 
-        self.losscao, self.weight_real, self.weight_fake = self.cao(pred_real0_ViT, pre_fake0_ViT, self.loss_D_real_ViT, self.loss_D_fake_ViT)
-        
-        return self.losscao* lambda_D_ViT
+        pred_real0_ViT = self.netD_ViT(real_B0_tokens)
+        pred_real1_ViT = self.netD_ViT(real_B1_tokens)
+        self.loss_D_real_ViT = (self.criterionGAN(pred_real0_ViT, True).mean() + self.criterionGAN(pred_real1_ViT, True).mean()) * 0.5 * lambda_D_ViT
+
+        self.loss_D_ViT = (self.loss_D_fake_ViT + self.loss_D_real_ViT) * 0.5
+
+
+        return self.loss_D_ViT
 
     def compute_E_loss(self):
         """计算判别器 E 的损失"""
         
+        print(f'resl_A_noisy: {self.real_A_noisy.shape} \n fake_B0: {self.fake_B0.shape}')
         XtXt_1 = torch.cat([self.real_A_noisy, self.fake_B0.detach()], dim=1)
         XtXt_2 = torch.cat([self.real_A_noisy2, self.fake_B1.detach()], dim=1)
         temp = torch.logsumexp(self.netE(XtXt_1, self.time, XtXt_2).reshape(-1), dim=0).mean()
@@ -528,28 +607,12 @@ class RomaUnsbModel(BaseModel):
         
     def compute_G_loss(self):
         """计算生成器的 GAN 损失"""
-        if self.opt.lambda_ctn > 0.0:
-            # 生成图像的CTN光流图
-            self.f_content = self.ctn(self.weight_fake)
-            
-            # 变换后的图片
-            self.warped_real_A_noisy2 = warp(self.real_A_noisy, self.f_content)
-            self.warped_fake_B0       = warp(self.fake_B0,self.f_content)
-            
-            # 经过第二次生成器
-            self.warped_fake_B0_2 = self.netG(self.warped_real_A_noisy2, self.time, self.z_in)   
-            
-            warped_fake_B0_2=self.warped_fake_B0_2
-            warped_fake_B0=self.warped_fake_B0
-            # 计算L2损失
-            self.ctn_loss = F.mse_loss(warped_fake_B0_2, warped_fake_B0)    
-
+        
         if self.opt.lambda_GAN > 0.0:
             pred_fake = self.netD_ViT(self.mutil_fake_B0_tokens[0])
-            self.loss_G_GAN = self.criterionGAN(pred_fake, True).mean() 
+            self.loss_G_GAN = self.criterionGAN(pred_fake, True).mean() * self.opt.lambda_GAN
         else:
             self.loss_G_GAN = 0.0
-            
         self.loss_SB = 0
         if self.opt.lambda_SB > 0.0:
             XtXt_1 = torch.cat([self.real_A_noisy, self.fake_B0], dim=1)
@@ -558,9 +621,9 @@ class RomaUnsbModel(BaseModel):
             bs = self.opt.batch_size
 
             # eq.9
-            ET_XY    = self.netE(XtXt_1, self.time, XtXt_1).mean() - self.netE(XtXt_1, self.time, XtXt_2).mean()
+            ET_XY    = self.netE(XtXt_1, self.time, XtXt_1).mean() - torch.logsumexp(self.netE(XtXt_1, self.time, XtXt_2).reshape(-1), dim=0)
             self.loss_SB = -(self.opt.num_timesteps - self.time[0]) / self.opt.num_timesteps * self.opt.tau * ET_XY
-            self.loss_SB += torch.mean((self.real_A_noisy - self.fake_B0) ** 2)
+            self.loss_SB += self.opt.tau * torch.mean((self.real_A_noisy - self.fake_B0) ** 2)
         
         if self.opt.lambda_global > 0.0:
             loss_global = self.calculate_similarity(self.real_A0, self.fake_B0) + self.calculate_similarity(self.real_A1, self.fake_B1)
@@ -568,10 +631,12 @@ class RomaUnsbModel(BaseModel):
         else:
             loss_global = 0.0
         
-        self.loss_G = self.opt.lambda_GAN * self.loss_G_GAN + \
-                      self.opt.lambda_SB * self.loss_SB + \
-                      self.opt.lambda_ctn * self.ctn_loss + \
-                      loss_global * self.opt.lambda_global
+        self.l2_loss = 0.0
+        if self.opt.lambda_l2 > 0.0:
+           wapped_fake_B = warp(self.fake_B0, self.f_content)  # use updated self.f_content
+           self.l2_loss = F.mse_loss(self.warped_fake_B0_2, wapped_fake_B)  # complete the loss calculation
+        
+        self.loss_G = self.loss_G_GAN + self.opt.lambda_SB * self.loss_SB + self.opt.lambda_ctn * self.l2_loss + loss_global * self.opt.lambda_global
         return self.loss_G
     
     def calculate_attention_loss(self):

roma.py

@@ -0,0 +1,301 @@
+import numpy as np
+import torch
+from .base_model import BaseModel
+from . import networks
+from .patchnce import PatchNCELoss
+import util.util as util
+import timm
+import time
+import torch.nn.functional as F
+import sys
+from functools import partial
+import torch.nn as nn
+import math
+
+from torchvision.transforms import transforms as tfs
+
+class ROMAModel(BaseModel):
+    
+    @staticmethod
+    def modify_commandline_options(parser, is_train=True):
+        """  Configures options specific for CUT model
+        """
+        parser.add_argument('--adj_size_list', type=list, default=[2, 4, 6, 8, 12], help='different scales of perception field')
+        parser.add_argument('--lambda_mlp', type=float, default=1.0, help='weight of lr for discriminator')
+        parser.add_argument('--lambda_motion', type=float, default=1.0, help='weight for Temporal Consistency')
+        parser.add_argument('--lambda_D_ViT', type=float, default=1.0, help='weight for discriminator')
+        parser.add_argument('--lambda_GAN', type=float, default=1.0, help='weight for GAN loss: GAN(G(X))')
+        parser.add_argument('--lambda_global', type=float, default=1.0, help='weight for Global Structural Consistency')
+        parser.add_argument('--lambda_spatial', type=float, default=1.0, help='weight for Local Structural Consistency')
+        parser.add_argument('--atten_layers', type=str, default='1,3,5', help='compute Cross-Similarity on which layers')
+        parser.add_argument('--local_nums', type=int, default=256)
+        parser.add_argument('--which_D_layer', type=int, default=-1)
+        parser.add_argument('--side_length', type=int, default=7)
+
+        parser.set_defaults(pool_size=0) 
+
+        opt, _ = parser.parse_known_args()
+
+        return parser
+
+    def __init__(self, opt):
+        BaseModel.__init__(self, opt)
+
+
+        self.loss_names = ['G_GAN_ViT', 'D_real_ViT', 'D_fake_ViT', 'global', 'spatial', 'motion']
+        self.visual_names = ['real_A0', 'real_A1', 'fake_B0', 'fake_B1', 'real_B0', 'real_B1']
+        self.atten_layers = [int(i) for i in self.opt.atten_layers.split(',')]
+
+
+        if self.isTrain:
+            self.model_names = ['G', 'D_ViT']
+        else:  # during test time, only load G
+            self.model_names = ['G']
+
+
+        # define networks (both generator and discriminator)
+        self.netG = networks.define_G(opt.input_nc, opt.output_nc, opt.ngf, opt.netG, opt.normG, not opt.no_dropout, opt.init_type, opt.init_gain, opt.no_antialias, opt.no_antialias_up, self.gpu_ids, opt)
+
+
+        if self.isTrain:
+
+            self.netD_ViT = networks.MLPDiscriminator().to(self.device)
+            self.netPreViT = timm.create_model("vit_base_patch16_384",pretrained=True).to(self.device)
+            
+
+            self.norm = F.softmax
+
+            self.resize = tfs.Resize(size=(384,384))
+
+            self.criterionGAN = networks.GANLoss(opt.gan_mode).to(self.device)
+            self.criterionNCE = []
+
+            for atten_layer in self.atten_layers:
+                self.criterionNCE.append(PatchNCELoss(opt).to(self.device))
+
+            self.criterionL1 = torch.nn.L1Loss().to(self.device)
+            self.optimizer_G = torch.optim.Adam(self.netG.parameters(), lr=opt.lr, betas=(opt.beta1, opt.beta2))
+            self.optimizer_D_ViT = torch.optim.Adam(self.netD_ViT.parameters(), lr=opt.lr * opt.lambda_mlp, betas=(opt.beta1, opt.beta2))
+            self.optimizers.append(self.optimizer_G)
+            self.optimizers.append(self.optimizer_D_ViT)
+
+    def data_dependent_initialize(self, data):
+        """
+        The feature network netF is defined in terms of the shape of the intermediate, extracted
+        features of the encoder portion of netG. Because of this, the weights of netF are
+        initialized at the first feedforward pass with some input images.
+        Please also see PatchSampleF.create_mlp(), which is called at the first forward() call.
+        """
+        pass
+   
+
+    def optimize_parameters(self):
+        # forward
+        self.forward()
+
+        # update D
+        self.set_requires_grad(self.netD_ViT, True)
+        self.optimizer_D_ViT.zero_grad()
+        self.loss_D = self.compute_D_loss()
+        self.loss_D.backward()
+        self.optimizer_D_ViT.step()
+
+        # update G
+        self.set_requires_grad(self.netD_ViT, False)
+        self.optimizer_G.zero_grad()
+        self.loss_G = self.compute_G_loss()
+        self.loss_G.backward()
+        self.optimizer_G.step()
+
+    def set_input(self, input):
+        """Unpack input data from the dataloader and perform necessary pre-processing steps.
+        Parameters:
+            input (dict): include the data itself and its metadata information.
+        The option 'direction' can be used to swap domain A and domain B.
+        """
+        AtoB = self.opt.direction == 'AtoB'
+        self.real_A0 = input['A0' if AtoB else 'B0'].to(self.device)
+        self.real_A1 = input['A1' if AtoB else 'B1'].to(self.device)
+        self.real_B0 = input['B0' if AtoB else 'A0'].to(self.device)
+        self.real_B1 = input['B1' if AtoB else 'A1'].to(self.device)
+        self.image_paths = input['A_paths' if AtoB else 'B_paths']
+
+    def forward(self):
+        """Run forward pass; called by both functions <optimize_parameters> and <test>."""
+        self.fake_B0 = self.netG(self.real_A0) 
+        self.fake_B1 = self.netG(self.real_A1)
+        
+        if self.opt.isTrain:
+            real_A0 = self.real_A0
+            real_A1 = self.real_A1
+            real_B0 = self.real_B0
+            real_B1 = self.real_B1
+            fake_B0 = self.fake_B0
+            fake_B1 = self.fake_B1
+            self.real_A0_resize = self.resize(real_A0)
+            self.real_A1_resize = self.resize(real_A1)
+            real_B0 = self.resize(real_B0)
+            real_B1 = self.resize(real_B1)
+            self.fake_B0_resize = self.resize(fake_B0)
+            self.fake_B1_resize = self.resize(fake_B1)
+            self.mutil_real_A0_tokens = self.netPreViT(self.real_A0_resize, self.atten_layers, get_tokens=True)
+            self.mutil_real_A1_tokens = self.netPreViT(self.real_A1_resize, self.atten_layers, get_tokens=True)
+            self.mutil_real_B0_tokens = self.netPreViT(real_B0, self.atten_layers, get_tokens=True)
+            self.mutil_real_B1_tokens = self.netPreViT(real_B1, self.atten_layers, get_tokens=True)
+            self.mutil_fake_B0_tokens = self.netPreViT(self.fake_B0_resize, self.atten_layers, get_tokens=True)
+            self.mutil_fake_B1_tokens = self.netPreViT(self.fake_B1_resize, self.atten_layers, get_tokens=True)
+
+    def tokens_concat(self, origin_tokens, adjacent_size):
+        adj_size = adjacent_size
+        B, token_num, C = origin_tokens.shape[0], origin_tokens.shape[1], origin_tokens.shape[2]
+        S = int(math.sqrt(token_num))
+        if S * S != token_num:
+            print('Error! Not a square!')
+        token_map = origin_tokens.clone().reshape(B,S,S,C)
+        cut_patch_list = []
+        for i in range(0, S, adj_size):
+            for j in range(0, S, adj_size):
+                i_left = i
+                i_right = i + adj_size + 1 if i + adj_size <= S else S + 1
+                j_left = j
+                j_right = j + adj_size if j + adj_size <= S else S + 1
+
+                cut_patch = token_map[:, i_left:i_right, j_left: j_right, :]
+                cut_patch= cut_patch.reshape(B,-1,C)
+                cut_patch = torch.mean(cut_patch, dim=1, keepdim=True)
+                cut_patch_list.append(cut_patch)
+
+
+        result = torch.cat(cut_patch_list,dim=1)
+        return result
+
+
+    def cat_results(self, origin_tokens, adj_size_list):
+        res_list = [origin_tokens]
+        for ad_s in adj_size_list:
+            cat_result = self.tokens_concat(origin_tokens, ad_s)
+            res_list.append(cat_result)
+        
+        result = torch.cat(res_list, dim=1)
+
+        return result
+
+
+
+    def compute_D_loss(self):
+        """Calculate GAN loss for the discriminator"""
+
+
+        lambda_D_ViT = self.opt.lambda_D_ViT
+        fake_B0_tokens = self.mutil_fake_B0_tokens[self.opt.which_D_layer].detach()
+        fake_B1_tokens = self.mutil_fake_B1_tokens[self.opt.which_D_layer].detach()
+
+        real_B0_tokens = self.mutil_real_B0_tokens[self.opt.which_D_layer]
+        real_B1_tokens = self.mutil_real_B1_tokens[self.opt.which_D_layer]
+
+
+        fake_B0_tokens = self.cat_results(fake_B0_tokens, self.opt.adj_size_list)
+        fake_B1_tokens = self.cat_results(fake_B1_tokens, self.opt.adj_size_list)
+
+
+
+        real_B0_tokens = self.cat_results(real_B0_tokens, self.opt.adj_size_list)
+        real_B1_tokens = self.cat_results(real_B1_tokens, self.opt.adj_size_list)
+
+        pre_fake0_ViT = self.netD_ViT(fake_B0_tokens)
+        pre_fake1_ViT = self.netD_ViT(fake_B1_tokens)
+
+        self.loss_D_fake_ViT = (self.criterionGAN(pre_fake0_ViT, False).mean() + self.criterionGAN(pre_fake1_ViT, False).mean()) * 0.5 * lambda_D_ViT
+
+        pred_real0_ViT = self.netD_ViT(real_B0_tokens)
+        pred_real1_ViT = self.netD_ViT(real_B1_tokens)
+        self.loss_D_real_ViT = (self.criterionGAN(pred_real0_ViT, True).mean() + self.criterionGAN(pred_real1_ViT, True).mean()) * 0.5 * lambda_D_ViT
+
+        self.loss_D_ViT = (self.loss_D_fake_ViT + self.loss_D_real_ViT) * 0.5
+
+
+        return self.loss_D_ViT
+
+    def compute_G_loss(self):
+
+        if self.opt.lambda_GAN > 0.0:
+
+            fake_B0_tokens = self.mutil_fake_B0_tokens[self.opt.which_D_layer]
+            fake_B1_tokens = self.mutil_fake_B1_tokens[self.opt.which_D_layer]
+            fake_B0_tokens = self.cat_results(fake_B0_tokens, self.opt.adj_size_list)
+            fake_B1_tokens = self.cat_results(fake_B1_tokens, self.opt.adj_size_list)
+            pred_fake0_ViT = self.netD_ViT(fake_B0_tokens)
+            pred_fake1_ViT = self.netD_ViT(fake_B1_tokens)
+            self.loss_G_GAN_ViT = (self.criterionGAN(pred_fake0_ViT, True) + self.criterionGAN(pred_fake1_ViT, True)) * 0.5 * self.opt.lambda_GAN
+        else:
+            self.loss_G_GAN_ViT = 0.0
+
+        if self.opt.lambda_global > 0.0 or self.opt.lambda_spatial > 0.0:
+            self.loss_global, self.loss_spatial = self.calculate_attention_loss()
+        else:
+            self.loss_global, self.loss_spatial = 0.0, 0.0
+
+        if self.opt.lambda_motion > 0.0:
+            self.loss_motion = 0.0
+            for real_A0_tokens, real_A1_tokens, fake_B0_tokens, fake_B1_tokens in zip(self.mutil_real_A0_tokens, self.mutil_real_A1_tokens, self.mutil_fake_B0_tokens, self.mutil_fake_B1_tokens):
+                A0_B1 = real_A0_tokens.bmm(fake_B1_tokens.permute(0,2,1))
+                B0_A1 = fake_B0_tokens.bmm(real_A1_tokens.permute(0,2,1))
+                cos_dis_global =  F.cosine_similarity(A0_B1, B0_A1, dim=-1)
+                self.loss_motion += self.criterionL1(torch.ones_like(cos_dis_global), cos_dis_global).mean()
+        else:
+            self.loss_motion = 0.0
+
+        self.loss_G = self.loss_G_GAN_ViT + self.loss_global + self.loss_spatial + self.loss_motion
+        return self.loss_G
+
+    def calculate_attention_loss(self):
+        n_layers = len(self.atten_layers)
+        mutil_real_A0_tokens = self.mutil_real_A0_tokens
+        mutil_real_A1_tokens = self.mutil_real_A1_tokens
+        mutil_fake_B0_tokens = self.mutil_fake_B0_tokens
+        mutil_fake_B1_tokens = self.mutil_fake_B1_tokens
+
+            
+        if self.opt.lambda_global > 0.0:
+            loss_global = self.calculate_similarity(mutil_real_A0_tokens, mutil_fake_B0_tokens) + self.calculate_similarity(mutil_real_A1_tokens, mutil_fake_B1_tokens)
+            loss_global *= 0.5
+
+        else:
+            loss_global = 0.0
+
+        if self.opt.lambda_spatial > 0.0:
+            loss_spatial = 0.0
+            local_nums = self.opt.local_nums
+            tokens_cnt = 576
+            local_id = np.random.permutation(tokens_cnt)
+            local_id = local_id[:int(min(local_nums, tokens_cnt))]
+
+            mutil_real_A0_local_tokens = self.netPreViT(self.real_A0_resize, self.atten_layers, get_tokens=True, local_id=local_id, side_length = self.opt.side_length)
+            mutil_real_A1_local_tokens = self.netPreViT(self.real_A1_resize, self.atten_layers, get_tokens=True, local_id=local_id, side_length = self.opt.side_length)
+
+            mutil_fake_B0_local_tokens = self.netPreViT(self.fake_B0_resize, self.atten_layers, get_tokens=True, local_id=local_id, side_length = self.opt.side_length)
+            mutil_fake_B1_local_tokens = self.netPreViT(self.fake_B1_resize, self.atten_layers, get_tokens=True, local_id=local_id, side_length = self.opt.side_length)
+
+            loss_spatial = self.calculate_similarity(mutil_real_A0_local_tokens, mutil_fake_B0_local_tokens) + self.calculate_similarity(mutil_real_A1_local_tokens, mutil_fake_B1_local_tokens)
+            loss_spatial *= 0.5
+        
+        else:
+            loss_spatial = 0.0
+
+            
+
+        return loss_global * self.opt.lambda_global, loss_spatial * self.opt.lambda_spatial
+
+    def calculate_similarity(self, mutil_src_tokens, mutil_tgt_tokens):
+        loss = 0.0
+        n_layers = len(self.atten_layers)
+
+        for src_tokens, tgt_tokens in zip(mutil_src_tokens, mutil_tgt_tokens):
+
+            src_tgt = src_tokens.bmm(tgt_tokens.permute(0,2,1))
+            tgt_src = tgt_tokens.bmm(src_tokens.permute(0,2,1))
+            cos_dis_global =  F.cosine_similarity(src_tgt, tgt_src, dim=-1)
+            loss += self.criterionL1(torch.ones_like(cos_dis_global), cos_dis_global).mean()
+
+        loss = loss / n_layers
+        return loss

scripts/train.sh

@@ -7,22 +7,27 @@
 
 python train.py \
     --dataroot /home/openxs/kunyu/datasets/InfraredCity-Lite/Double/Moitor \
-    --name ROMA_UNSB_003 \
+    --name ROMA_UNSB_001 \
     --dataset_mode unaligned_double \
     --no_flip \
     --display_env ROMA \
     --model roma_unsb \
-    --lambda_GAN 1.0 \
+    --lambda_GAN 8.0 \
     --lambda_NCE 8.0 \
-    --lambda_SB 1.0 \
+    --lambda_SB 0.1 \
     --lambda_ctn 1.0 \
     --lambda_inc 1.0 \
     --lr 0.00001 \
     --gpu_id 0 \
     --nce_idt False \
+    --nce_layers 0,4,8,12,16 \
     --netF mlp_sample \
-    --flip_equivariance True \
-    --eta_ratio 0.4 \
+    --netF_nc 256 \
+    --nce_T 0.07 \
+    --lmda_1 0.1 \
+    --num_patches 256 \
+    --flip_equivariance False \
+    --eta_ratio 0.1 \
     --tau 0.01 \
-    --num_timesteps 4 \
+    --num_timesteps 10 \
     --input_nc 3

train.py

@@ -44,7 +44,6 @@ if __name__ == '__main__':
                 model.setup(opt)               # regular setup: load and print networks; create schedulers
                 model.parallelize()
             model.set_input(data)  # unpack data from dataset and apply preprocessing
-            #print('Call opt paras')
             model.optimize_parameters()   # calculate loss functions, get gradients, update network weights
             if len(opt.gpu_ids) > 0:
                 torch.cuda.synchronize()