清理代码,删除跳连接部分
This commit is contained in:
@@ -45,6 +45,15 @@ def denormalize(tensor):
|
|||||||
# [0, 1] -> [0, 255]
|
# [0, 1] -> [0, 255]
|
||||||
tensor = tensor * 255
|
tensor = tensor * 255
|
||||||
return tensor.clamp(0, 255)
|
return tensor.clamp(0, 255)
|
||||||
|
# return tensor
|
||||||
|
|
||||||
|
def minmax_denormalize(tensor):
|
||||||
|
tensor_min = tensor.min()
|
||||||
|
tensor_max = tensor.max()
|
||||||
|
tensor = (tensor - tensor_min) / (tensor_max - tensor_min)
|
||||||
|
# tensor = tensor*2-1
|
||||||
|
tensor = tensor*255
|
||||||
|
return tensor.clamp(0, 255)
|
||||||
|
|
||||||
|
|
||||||
def calculate_metrics(pred, target, debug=False):
|
def calculate_metrics(pred, target, debug=False):
|
||||||
@@ -134,6 +143,10 @@ def save_comparison_figure(input_frames, target_frame, pred_frame, save_path,
|
|||||||
ax.set_title('Predicted')
|
ax.set_title('Predicted')
|
||||||
ax.axis('off')
|
ax.axis('off')
|
||||||
|
|
||||||
|
#debug print
|
||||||
|
print(target_frame)
|
||||||
|
print(pred_frame)
|
||||||
|
|
||||||
# debug print - 改进为更有信息量的输出
|
# debug print - 改进为更有信息量的输出
|
||||||
if isinstance(pred_frame, np.ndarray):
|
if isinstance(pred_frame, np.ndarray):
|
||||||
print(f"[DEBUG IMAGE] Pred frame shape: {pred_frame.shape}, range: [{pred_frame.min():.2f}, {pred_frame.max():.2f}], mean: {pred_frame.mean():.2f}")
|
print(f"[DEBUG IMAGE] Pred frame shape: {pred_frame.shape}, range: [{pred_frame.min():.2f}, {pred_frame.max():.2f}], mean: {pred_frame.mean():.2f}")
|
||||||
@@ -161,8 +174,8 @@ def evaluate_model(model, data_loader, device, args):
|
|||||||
metrics_dict: 包含所有指标的字典
|
metrics_dict: 包含所有指标的字典
|
||||||
sample_results: 示例结果用于可视化
|
sample_results: 示例结果用于可视化
|
||||||
"""
|
"""
|
||||||
# model.eval()
|
model.eval()
|
||||||
model.train() # 临时使用训练模式
|
# model.train() # 临时使用训练模式
|
||||||
|
|
||||||
# 初始化指标累加器
|
# 初始化指标累加器
|
||||||
total_mse = 0.0
|
total_mse = 0.0
|
||||||
@@ -183,10 +196,11 @@ def evaluate_model(model, data_loader, device, args):
|
|||||||
target_frames = target_frames.to(device, non_blocking=True)
|
target_frames = target_frames.to(device, non_blocking=True)
|
||||||
|
|
||||||
# 前向传播
|
# 前向传播
|
||||||
pred_frames, _ = model(input_frames)
|
pred_frames = model(input_frames)
|
||||||
|
|
||||||
# 反归一化用于指标计算
|
# 反归一化用于指标计算
|
||||||
pred_denorm = denormalize(pred_frames) # [B, 1, H, W]
|
# pred_denorm = minmax_denormalize(pred_frames) # [B, 1, H, W]
|
||||||
|
pred_denorm = denormalize(pred_frames)
|
||||||
target_denorm = denormalize(target_frames) # [B, 1, H, W]
|
target_denorm = denormalize(target_frames) # [B, 1, H, W]
|
||||||
|
|
||||||
batch_size = input_frames.size(0)
|
batch_size = input_frames.size(0)
|
||||||
@@ -309,8 +323,6 @@ def main(args):
|
|||||||
print(f"创建模型: {args.model}")
|
print(f"创建模型: {args.model}")
|
||||||
model_kwargs = {
|
model_kwargs = {
|
||||||
'num_frames': args.num_frames,
|
'num_frames': args.num_frames,
|
||||||
'use_representation_head': args.use_representation_head,
|
|
||||||
'representation_dim': args.representation_dim,
|
|
||||||
}
|
}
|
||||||
|
|
||||||
if args.model == 'SwiftFormerTemporal_XS':
|
if args.model == 'SwiftFormerTemporal_XS':
|
||||||
@@ -335,10 +347,10 @@ def main(args):
|
|||||||
except (pickle.UnpicklingError, TypeError) as e:
|
except (pickle.UnpicklingError, TypeError) as e:
|
||||||
print(f"使用weights_only=False加载失败: {e}")
|
print(f"使用weights_only=False加载失败: {e}")
|
||||||
print("尝试使用torch.serialization.add_safe_globals...")
|
print("尝试使用torch.serialization.add_safe_globals...")
|
||||||
from argparse import Namespace
|
# from argparse import Namespace
|
||||||
# 添加安全全局变量
|
# # 添加安全全局变量
|
||||||
torch.serialization.add_safe_globals([Namespace])
|
# torch.serialization.add_safe_globals([Namespace])
|
||||||
checkpoint = torch.load(args.resume, map_location='cpu')
|
# checkpoint = torch.load(args.resume, map_location='cpu')
|
||||||
|
|
||||||
# 处理状态字典(可能包含'module.'前缀)
|
# 处理状态字典(可能包含'module.'前缀)
|
||||||
if 'model' in checkpoint:
|
if 'model' in checkpoint:
|
||||||
@@ -462,10 +474,6 @@ def get_args_parser():
|
|||||||
# 模型参数
|
# 模型参数
|
||||||
parser.add_argument('--model', default='SwiftFormerTemporal_XS', type=str, metavar='MODEL',
|
parser.add_argument('--model', default='SwiftFormerTemporal_XS', type=str, metavar='MODEL',
|
||||||
help='要评估的模型名称')
|
help='要评估的模型名称')
|
||||||
parser.add_argument('--use-representation-head', action='store_true',
|
|
||||||
help='使用表示头进行姿态/速度预测')
|
|
||||||
parser.add_argument('--representation-dim', default=128, type=int,
|
|
||||||
help='表示向量的维度')
|
|
||||||
|
|
||||||
# 评估参数
|
# 评估参数
|
||||||
parser.add_argument('--batch-size', default=16, type=int,
|
parser.add_argument('--batch-size', default=16, type=int,
|
||||||
|
|||||||
@@ -49,11 +49,6 @@ def get_args_parser():
|
|||||||
# Model parameters
|
# Model parameters
|
||||||
parser.add_argument('--model', default='SwiftFormerTemporal_XS', type=str, metavar='MODEL',
|
parser.add_argument('--model', default='SwiftFormerTemporal_XS', type=str, metavar='MODEL',
|
||||||
help='Name of model to train')
|
help='Name of model to train')
|
||||||
parser.add_argument('--use-representation-head', action='store_true',
|
|
||||||
help='Use representation head for pose/velocity prediction')
|
|
||||||
parser.add_argument('--representation-dim', default=128, type=int,
|
|
||||||
help='Dimension of representation vector')
|
|
||||||
parser.add_argument('--use-skip', default=False, type=bool, help='using skip connections')
|
|
||||||
|
|
||||||
# Training parameters
|
# Training parameters
|
||||||
parser.add_argument('--batch-size', default=32, type=int)
|
parser.add_argument('--batch-size', default=32, type=int)
|
||||||
@@ -207,9 +202,6 @@ def main(args):
|
|||||||
print(f"Creating model: {args.model}")
|
print(f"Creating model: {args.model}")
|
||||||
model_kwargs = {
|
model_kwargs = {
|
||||||
'num_frames': args.num_frames,
|
'num_frames': args.num_frames,
|
||||||
'use_representation_head': args.use_representation_head,
|
|
||||||
'representation_dim': args.representation_dim,
|
|
||||||
'use_skip': args.use_skip,
|
|
||||||
}
|
}
|
||||||
|
|
||||||
if args.model == 'SwiftFormerTemporal_XS':
|
if args.model == 'SwiftFormerTemporal_XS':
|
||||||
@@ -258,7 +250,7 @@ def main(args):
|
|||||||
super().__init__()
|
super().__init__()
|
||||||
self.mse = nn.MSELoss()
|
self.mse = nn.MSELoss()
|
||||||
|
|
||||||
def forward(self, pred_frame, target_frame, representations=None, temporal_indices=None):
|
def forward(self, pred_frame, target_frame, temporal_indices=None):
|
||||||
loss = self.mse(pred_frame, target_frame)
|
loss = self.mse(pred_frame, target_frame)
|
||||||
loss_dict = {'mse': loss}
|
loss_dict = {'mse': loss}
|
||||||
return loss, loss_dict
|
return loss, loss_dict
|
||||||
@@ -386,10 +378,10 @@ def train_one_epoch(model, criterion, data_loader, optimizer, device, epoch, los
|
|||||||
|
|
||||||
# Forward pass
|
# Forward pass
|
||||||
with torch.amp.autocast(device_type='cuda'):
|
with torch.amp.autocast(device_type='cuda'):
|
||||||
pred_frames, representations = model(input_frames)
|
pred_frames = model(input_frames)
|
||||||
loss, loss_dict = criterion(
|
loss, loss_dict = criterion(
|
||||||
pred_frames, target_frames,
|
pred_frames, target_frames,
|
||||||
representations, temporal_indices
|
temporal_indices
|
||||||
)
|
)
|
||||||
|
|
||||||
loss_value = loss.item()
|
loss_value = loss.item()
|
||||||
@@ -452,7 +444,7 @@ def train_one_epoch(model, criterion, data_loader, optimizer, device, epoch, los
|
|||||||
if args is not None and getattr(args, 'log_images', False) and global_step % getattr(args, 'image_log_freq', 100) == 0:
|
if args is not None and getattr(args, 'log_images', False) and global_step % getattr(args, 'image_log_freq', 100) == 0:
|
||||||
with torch.no_grad():
|
with torch.no_grad():
|
||||||
# Take first sample from batch for visualization
|
# Take first sample from batch for visualization
|
||||||
pred_vis, _ = model(input_frames[:1])
|
pred_vis = model(input_frames[:1])
|
||||||
# Convert to appropriate format for TensorBoard
|
# Convert to appropriate format for TensorBoard
|
||||||
# Assuming frames are in [B, C, H, W] format
|
# Assuming frames are in [B, C, H, W] format
|
||||||
writer.add_images('train/input', input_frames[:1], global_step)
|
writer.add_images('train/input', input_frames[:1], global_step)
|
||||||
@@ -497,10 +489,10 @@ def evaluate(data_loader, model, criterion, device, writer=None, epoch=0):
|
|||||||
|
|
||||||
# Compute output
|
# Compute output
|
||||||
with torch.amp.autocast(device_type='cuda'):
|
with torch.amp.autocast(device_type='cuda'):
|
||||||
pred_frames, representations = model(input_frames)
|
pred_frames = model(input_frames)
|
||||||
loss, loss_dict = criterion(
|
loss, loss_dict = criterion(
|
||||||
pred_frames, target_frames,
|
pred_frames, target_frames,
|
||||||
representations, temporal_indices
|
temporal_indices
|
||||||
)
|
)
|
||||||
|
|
||||||
# 计算诊断指标
|
# 计算诊断指标
|
||||||
|
|||||||
@@ -93,159 +93,33 @@ class DecoderBlock(nn.Module):
|
|||||||
return x
|
return x
|
||||||
|
|
||||||
|
|
||||||
class DecoderBlockWithSkip(nn.Module):
|
|
||||||
"""Decoder block with skip connection support"""
|
|
||||||
def __init__(self, in_channels, out_channels, skip_channels=0, kernel_size=3, stride=2, padding=1, output_padding=1):
|
|
||||||
super().__init__()
|
|
||||||
# 总输入通道 = 输入通道 + skip通道
|
|
||||||
total_in_channels = in_channels + skip_channels
|
|
||||||
|
|
||||||
# 主路径:反卷积 + 两个卷积层
|
|
||||||
self.conv_transpose = nn.ConvTranspose2d(
|
|
||||||
total_in_channels, out_channels,
|
|
||||||
kernel_size=kernel_size,
|
|
||||||
stride=stride,
|
|
||||||
padding=padding,
|
|
||||||
output_padding=output_padding,
|
|
||||||
bias=True
|
|
||||||
)
|
|
||||||
self.conv1 = nn.Conv2d(out_channels, out_channels,
|
|
||||||
kernel_size=3, padding=1, bias=True)
|
|
||||||
self.conv2 = nn.Conv2d(out_channels, out_channels,
|
|
||||||
kernel_size=3, padding=1, bias=True)
|
|
||||||
|
|
||||||
# 残差路径:如果需要改变通道数或空间尺寸
|
|
||||||
self.shortcut = nn.Identity()
|
|
||||||
if total_in_channels != out_channels or stride != 1:
|
|
||||||
if stride == 1:
|
|
||||||
self.shortcut = nn.Conv2d(total_in_channels, out_channels,
|
|
||||||
kernel_size=1, bias=True)
|
|
||||||
else:
|
|
||||||
self.shortcut = nn.ConvTranspose2d(
|
|
||||||
total_in_channels, out_channels,
|
|
||||||
kernel_size=1,
|
|
||||||
stride=stride,
|
|
||||||
padding=0,
|
|
||||||
output_padding=output_padding,
|
|
||||||
bias=True
|
|
||||||
)
|
|
||||||
|
|
||||||
# 使用LeakyReLU避免死亡神经元
|
|
||||||
self.activation = nn.LeakyReLU(0.2, inplace=True)
|
|
||||||
|
|
||||||
# 初始化权重
|
|
||||||
self._init_weights()
|
|
||||||
|
|
||||||
def _init_weights(self):
|
|
||||||
# 初始化反卷积层
|
|
||||||
nn.init.kaiming_normal_(self.conv_transpose.weight, mode='fan_out', nonlinearity='leaky_relu')
|
|
||||||
if self.conv_transpose.bias is not None:
|
|
||||||
nn.init.constant_(self.conv_transpose.bias, 0)
|
|
||||||
|
|
||||||
# 初始化卷积层
|
|
||||||
nn.init.kaiming_normal_(self.conv1.weight, mode='fan_out', nonlinearity='leaky_relu')
|
|
||||||
if self.conv1.bias is not None:
|
|
||||||
nn.init.constant_(self.conv1.bias, 0)
|
|
||||||
|
|
||||||
nn.init.kaiming_normal_(self.conv2.weight, mode='fan_out', nonlinearity='leaky_relu')
|
|
||||||
if self.conv2.bias is not None:
|
|
||||||
nn.init.constant_(self.conv2.bias, 0)
|
|
||||||
|
|
||||||
# 初始化shortcut
|
|
||||||
if not isinstance(self.shortcut, nn.Identity):
|
|
||||||
if isinstance(self.shortcut, nn.Conv2d):
|
|
||||||
nn.init.kaiming_normal_(self.shortcut.weight, mode='fan_out', nonlinearity='leaky_relu')
|
|
||||||
elif isinstance(self.shortcut, nn.ConvTranspose2d):
|
|
||||||
nn.init.kaiming_normal_(self.shortcut.weight, mode='fan_out', nonlinearity='leaky_relu')
|
|
||||||
if self.shortcut.bias is not None:
|
|
||||||
nn.init.constant_(self.shortcut.bias, 0)
|
|
||||||
|
|
||||||
def forward(self, x, skip_feature=None):
|
|
||||||
# 如果有skip feature,将其与输入拼接
|
|
||||||
if skip_feature is not None:
|
|
||||||
# 确保skip特征的空间尺寸与x匹配
|
|
||||||
if skip_feature.shape[2:] != x.shape[2:]:
|
|
||||||
# 使用双线性插值进行上采样或下采样
|
|
||||||
skip_feature = torch.nn.functional.interpolate(
|
|
||||||
skip_feature,
|
|
||||||
size=x.shape[2:],
|
|
||||||
mode='bilinear',
|
|
||||||
align_corners=False
|
|
||||||
)
|
|
||||||
x = torch.cat([x, skip_feature], dim=1)
|
|
||||||
|
|
||||||
identity = self.shortcut(x)
|
|
||||||
|
|
||||||
# 主路径
|
|
||||||
x = self.conv_transpose(x)
|
|
||||||
x = self.activation(x)
|
|
||||||
|
|
||||||
x = self.conv1(x)
|
|
||||||
x = self.activation(x)
|
|
||||||
|
|
||||||
x = self.conv2(x)
|
|
||||||
|
|
||||||
# 残差连接
|
|
||||||
x = x + identity
|
|
||||||
x = self.activation(x)
|
|
||||||
return x
|
|
||||||
|
|
||||||
|
|
||||||
class FramePredictionDecoder(nn.Module):
|
class FramePredictionDecoder(nn.Module):
|
||||||
"""Improved decoder for frame prediction with better upsampling strategy"""
|
"""Improved decoder for frame prediction"""
|
||||||
def __init__(self, embed_dims, output_channels=1, use_skip=False):
|
def __init__(self, embed_dims, output_channels=1):
|
||||||
super().__init__()
|
super().__init__()
|
||||||
self.use_skip = use_skip
|
|
||||||
# Reverse the embed_dims for decoder
|
# Reverse the embed_dims for decoder
|
||||||
decoder_dims = embed_dims[::-1]
|
decoder_dims = embed_dims[::-1]
|
||||||
|
|
||||||
self.blocks = nn.ModuleList()
|
self.blocks = nn.ModuleList()
|
||||||
|
|
||||||
if use_skip:
|
# 使用普通的DecoderBlock,第一个block使用大步长
|
||||||
# 使用支持skip connections的block
|
self.blocks.append(DecoderBlock(
|
||||||
# 第一个block:从bottleneck到stage4,使用大步长stride=4,skip来自stage3
|
decoder_dims[0], decoder_dims[1],
|
||||||
self.blocks.append(DecoderBlockWithSkip(
|
kernel_size=3, stride=4, padding=1, output_padding=3 # 改为stride=4
|
||||||
decoder_dims[0], decoder_dims[1],
|
))
|
||||||
skip_channels=embed_dims[3], # stage3的通道数
|
self.blocks.append(DecoderBlock(
|
||||||
kernel_size=3, stride=4, padding=1, output_padding=3 # 改为stride=4
|
decoder_dims[1], decoder_dims[2],
|
||||||
))
|
kernel_size=3, stride=2, padding=1, output_padding=1
|
||||||
# 第二个block:stage4到stage3,stride=2,skip来自stage2
|
))
|
||||||
self.blocks.append(DecoderBlockWithSkip(
|
self.blocks.append(DecoderBlock(
|
||||||
decoder_dims[1], decoder_dims[2],
|
decoder_dims[2], decoder_dims[3],
|
||||||
skip_channels=embed_dims[2], # stage2的通道数
|
kernel_size=3, stride=2, padding=1, output_padding=1
|
||||||
kernel_size=3, stride=2, padding=1, output_padding=1
|
))
|
||||||
))
|
# 第四个block:增加到64通道
|
||||||
# 第三个block:stage3到stage2,stride=2,skip来自stage1
|
self.blocks.append(DecoderBlock(
|
||||||
self.blocks.append(DecoderBlockWithSkip(
|
decoder_dims[3], 64,
|
||||||
decoder_dims[2], decoder_dims[3],
|
kernel_size=3, stride=2, padding=1, output_padding=1
|
||||||
skip_channels=embed_dims[1], # stage1的通道数
|
))
|
||||||
kernel_size=3, stride=2, padding=1, output_padding=1
|
|
||||||
))
|
|
||||||
# 第四个block:stage2到stage1,stride=2,skip来自stage0
|
|
||||||
self.blocks.append(DecoderBlockWithSkip(
|
|
||||||
decoder_dims[3], 64, # 输出到64通道
|
|
||||||
skip_channels=embed_dims[0], # stage0的通道数
|
|
||||||
kernel_size=3, stride=2, padding=1, output_padding=1
|
|
||||||
))
|
|
||||||
else:
|
|
||||||
# 使用普通的DecoderBlock,第一个block使用大步长
|
|
||||||
self.blocks.append(DecoderBlock(
|
|
||||||
decoder_dims[0], decoder_dims[1],
|
|
||||||
kernel_size=3, stride=4, padding=1, output_padding=3 # 改为stride=4
|
|
||||||
))
|
|
||||||
self.blocks.append(DecoderBlock(
|
|
||||||
decoder_dims[1], decoder_dims[2],
|
|
||||||
kernel_size=3, stride=2, padding=1, output_padding=1
|
|
||||||
))
|
|
||||||
self.blocks.append(DecoderBlock(
|
|
||||||
decoder_dims[2], decoder_dims[3],
|
|
||||||
kernel_size=3, stride=2, padding=1, output_padding=1
|
|
||||||
))
|
|
||||||
# 第四个block:增加到64通道
|
|
||||||
self.blocks.append(DecoderBlock(
|
|
||||||
decoder_dims[3], 64,
|
|
||||||
kernel_size=3, stride=2, padding=1, output_padding=1
|
|
||||||
))
|
|
||||||
|
|
||||||
# 改进的最终输出层:不使用反卷积,只进行特征精炼
|
# 改进的最终输出层:不使用反卷积,只进行特征精炼
|
||||||
# 输入尺寸已经是目标尺寸,只需要调整通道数和进行特征融合
|
# 输入尺寸已经是目标尺寸,只需要调整通道数和进行特征融合
|
||||||
@@ -258,48 +132,14 @@ class FramePredictionDecoder(nn.Module):
|
|||||||
# 移除Tanh,让输出在任意范围,由损失函数和归一化处理
|
# 移除Tanh,让输出在任意范围,由损失函数和归一化处理
|
||||||
)
|
)
|
||||||
|
|
||||||
def forward(self, x, skip_features=None):
|
def forward(self, x):
|
||||||
"""
|
"""
|
||||||
Args:
|
Args:
|
||||||
x: input tensor of shape [B, embed_dims[-1], H/32, W/32]
|
x: input tensor of shape [B, embed_dims[-1], H/32, W/32]
|
||||||
skip_features: list of encoder features from stages [stage3, stage2, stage1, stage0]
|
|
||||||
each of shape [B, C, H', W'] where C matches encoder dims
|
|
||||||
"""
|
"""
|
||||||
if self.use_skip:
|
# 不使用skip connections
|
||||||
if skip_features is None:
|
for i in range(4):
|
||||||
raise ValueError("skip_features must be provided when use_skip=True")
|
x = self.blocks[i](x)
|
||||||
|
|
||||||
# 确保有4个skip features
|
|
||||||
assert len(skip_features) == 4, f"Need 4 skip features, got {len(skip_features)}"
|
|
||||||
|
|
||||||
# 反转顺序以匹配解码器:stage3, stage2, stage1, stage0
|
|
||||||
skip_features = skip_features[::-1]
|
|
||||||
|
|
||||||
# 调整skip特征的尺寸以匹配新的上采样策略
|
|
||||||
adjusted_skip_features = []
|
|
||||||
for i, skip in enumerate(skip_features):
|
|
||||||
if skip is not None:
|
|
||||||
# 计算目标尺寸:4, 2, 2, 2倍上采样
|
|
||||||
upsample_factors = [4, 2, 2, 2]
|
|
||||||
target_height = x.shape[2] * upsample_factors[i]
|
|
||||||
target_width = x.shape[3] * upsample_factors[i]
|
|
||||||
|
|
||||||
if skip.shape[2:] != (target_height, target_width):
|
|
||||||
skip = torch.nn.functional.interpolate(
|
|
||||||
skip,
|
|
||||||
size=(target_height, target_width),
|
|
||||||
mode='bilinear',
|
|
||||||
align_corners=False
|
|
||||||
)
|
|
||||||
adjusted_skip_features.append(skip)
|
|
||||||
|
|
||||||
# 四个block使用skip connections
|
|
||||||
for i in range(4):
|
|
||||||
x = self.blocks[i](x, adjusted_skip_features[i])
|
|
||||||
else:
|
|
||||||
# 不使用skip connections
|
|
||||||
for i in range(4):
|
|
||||||
x = self.blocks[i](x)
|
|
||||||
|
|
||||||
# 最终输出层:只进行特征精炼,不上采样
|
# 最终输出层:只进行特征精炼,不上采样
|
||||||
x = self.final_block(x)
|
x = self.final_block(x)
|
||||||
@@ -316,9 +156,6 @@ class SwiftFormerTemporal(nn.Module):
|
|||||||
model_name='XS',
|
model_name='XS',
|
||||||
num_frames=3,
|
num_frames=3,
|
||||||
use_decoder=True,
|
use_decoder=True,
|
||||||
use_skip=True, # 新增:是否使用skip connections
|
|
||||||
use_representation_head=False,
|
|
||||||
representation_dim=128,
|
|
||||||
return_features=False,
|
return_features=False,
|
||||||
**kwargs):
|
**kwargs):
|
||||||
super().__init__()
|
super().__init__()
|
||||||
@@ -330,8 +167,6 @@ class SwiftFormerTemporal(nn.Module):
|
|||||||
# Store configuration
|
# Store configuration
|
||||||
self.num_frames = num_frames
|
self.num_frames = num_frames
|
||||||
self.use_decoder = use_decoder
|
self.use_decoder = use_decoder
|
||||||
self.use_skip = use_skip # 保存skip connections设置
|
|
||||||
self.use_representation_head = use_representation_head
|
|
||||||
self.return_features = return_features
|
self.return_features = return_features
|
||||||
|
|
||||||
# Modify stem to accept multiple frames (only Y channel)
|
# Modify stem to accept multiple frames (only Y channel)
|
||||||
@@ -365,22 +200,9 @@ class SwiftFormerTemporal(nn.Module):
|
|||||||
if use_decoder:
|
if use_decoder:
|
||||||
self.decoder = FramePredictionDecoder(
|
self.decoder = FramePredictionDecoder(
|
||||||
embed_dims,
|
embed_dims,
|
||||||
output_channels=1,
|
output_channels=1
|
||||||
use_skip=use_skip # 传递skip connections设置
|
|
||||||
)
|
)
|
||||||
|
|
||||||
# Representation head for pose/velocity prediction
|
|
||||||
if use_representation_head:
|
|
||||||
self.representation_head = nn.Sequential(
|
|
||||||
nn.AdaptiveAvgPool2d(1),
|
|
||||||
nn.Flatten(),
|
|
||||||
nn.Linear(embed_dims[-1], representation_dim),
|
|
||||||
nn.ReLU(),
|
|
||||||
nn.Linear(representation_dim, representation_dim)
|
|
||||||
)
|
|
||||||
else:
|
|
||||||
self.representation_head = None
|
|
||||||
|
|
||||||
self.apply(self._init_weights)
|
self.apply(self._init_weights)
|
||||||
|
|
||||||
def _init_weights(self, m):
|
def _init_weights(self, m):
|
||||||
@@ -400,7 +222,7 @@ class SwiftFormerTemporal(nn.Module):
|
|||||||
|
|
||||||
def forward_tokens(self, x):
|
def forward_tokens(self, x):
|
||||||
"""Forward through encoder network, return list of stage features if return_features else final output"""
|
"""Forward through encoder network, return list of stage features if return_features else final output"""
|
||||||
if self.return_features or self.use_skip:
|
if self.return_features:
|
||||||
features = []
|
features = []
|
||||||
stage_idx = 0
|
stage_idx = 0
|
||||||
for idx, block in enumerate(self.network):
|
for idx, block in enumerate(self.network):
|
||||||
@@ -423,59 +245,37 @@ class SwiftFormerTemporal(nn.Module):
|
|||||||
Returns:
|
Returns:
|
||||||
If return_features is False:
|
If return_features is False:
|
||||||
pred_frame: predicted frame [B, 1, H, W] (or None)
|
pred_frame: predicted frame [B, 1, H, W] (or None)
|
||||||
representation: optional representation vector [B, representation_dim] (or None)
|
|
||||||
If return_features is True:
|
If return_features is True:
|
||||||
pred_frame, representation, features (list of stage features)
|
pred_frame, features (list of stage features)
|
||||||
"""
|
"""
|
||||||
# Encode
|
# Encode
|
||||||
x = self.patch_embed(x)
|
x = self.patch_embed(x)
|
||||||
if self.return_features or self.use_skip:
|
if self.return_features:
|
||||||
x, features = self.forward_tokens(x)
|
x, features = self.forward_tokens(x)
|
||||||
else:
|
else:
|
||||||
x = self.forward_tokens(x)
|
x = self.forward_tokens(x)
|
||||||
x = self.norm(x)
|
x = self.norm(x)
|
||||||
|
|
||||||
# Get representation if needed
|
|
||||||
representation = None
|
|
||||||
if self.representation_head is not None:
|
|
||||||
representation = self.representation_head(x)
|
|
||||||
|
|
||||||
# Decode to frame
|
# Decode to frame
|
||||||
pred_frame = None
|
pred_frame = None
|
||||||
if self.use_decoder:
|
if self.use_decoder:
|
||||||
if self.use_skip:
|
pred_frame = self.decoder(x)
|
||||||
# 提取用于skip connections的特征
|
|
||||||
# features包含所有stage的输出,我们需要stage0, stage1, stage2, stage3
|
|
||||||
# 根据SwiftFormer结构,应该有4个stage特征
|
|
||||||
if len(features) >= 4:
|
|
||||||
# 取四个stage的特征:stage0, stage1, stage2, stage3
|
|
||||||
skip_features = [features[0], features[1], features[2], features[3]]
|
|
||||||
else:
|
|
||||||
# 如果特征不够,使用可用的特征
|
|
||||||
skip_features = features[:4]
|
|
||||||
# 如果特征仍然不够,使用None填充
|
|
||||||
while len(skip_features) < 4:
|
|
||||||
skip_features.append(None)
|
|
||||||
|
|
||||||
pred_frame = self.decoder(x, skip_features)
|
|
||||||
else:
|
|
||||||
pred_frame = self.decoder(x)
|
|
||||||
|
|
||||||
if self.return_features:
|
if self.return_features:
|
||||||
return pred_frame, representation, features
|
return pred_frame, features
|
||||||
else:
|
else:
|
||||||
return pred_frame, representation
|
return pred_frame
|
||||||
|
|
||||||
|
|
||||||
# Factory functions for different model sizes
|
# Factory functions for different model sizes
|
||||||
def SwiftFormerTemporal_XS(num_frames=3, use_skip=True, **kwargs):
|
def SwiftFormerTemporal_XS(num_frames=3, **kwargs):
|
||||||
return SwiftFormerTemporal('XS', num_frames=num_frames, use_skip=use_skip, **kwargs)
|
return SwiftFormerTemporal('XS', num_frames=num_frames, **kwargs)
|
||||||
|
|
||||||
def SwiftFormerTemporal_S(num_frames=3, use_skip=True, **kwargs):
|
def SwiftFormerTemporal_S(num_frames=3, **kwargs):
|
||||||
return SwiftFormerTemporal('S', num_frames=num_frames, use_skip=use_skip, **kwargs)
|
return SwiftFormerTemporal('S', num_frames=num_frames, **kwargs)
|
||||||
|
|
||||||
def SwiftFormerTemporal_L1(num_frames=3, use_skip=True, **kwargs):
|
def SwiftFormerTemporal_L1(num_frames=3, **kwargs):
|
||||||
return SwiftFormerTemporal('l1', num_frames=num_frames, use_skip=use_skip, **kwargs)
|
return SwiftFormerTemporal('l1', num_frames=num_frames, **kwargs)
|
||||||
|
|
||||||
def SwiftFormerTemporal_L3(num_frames=3, use_skip=True, **kwargs):
|
def SwiftFormerTemporal_L3(num_frames=3, **kwargs):
|
||||||
return SwiftFormerTemporal('l3', num_frames=num_frames, use_skip=use_skip, **kwargs)
|
return SwiftFormerTemporal('l3', num_frames=num_frames, **kwargs)
|
||||||
Reference in New Issue
Block a user