目标检测基础
目标检测与边界框
Program:
img = Image.open('/home/yuzhu/2.jpeg')
# bounding boxes
hiromi_bbox, michiko_bbox = [145, 25, 247, 150], [260, 50, 350, 160]
def bbox_to_rect(bbox, color):
return plt.Rectangle(xy=(bbox[0], bbox[1]), width=bbox[2]-bbox[0], height=bbox[3]-bbox[1], fill=False, edgecolor=color, linewidth=2)
fig = plt.imshow(img)
fig.axes.add_patch(bbox_to_rect(hiromi_bbox, 'blue'))
fig.axes.add_patch(bbox_to_rect(michiko_bbox, 'red'))
plt.show()
锚框
目标检测算法通常会在输入图像中采样大量的区域,然后判断这些区域中是否包含我们感兴趣的目标,并调整区域边缘从而更准确地预测目标的真实边界框,这里介绍的一种方法是以每个像素为中心生成多个大小和宽高比(aspect ratio)不同的边界框,这些边界框被称为锚框(anchor box)
生成多个锚框
交并比
# 以下函数已保存在d2lzh_pytorch包中方便以后使用
def assign_anchor(bb, anchor, jaccard_threshold=0.5):
"""
# 按照「9.4.1. 生成多个锚框」图9.3所讲为每个anchor分配真实的bb, anchor表示成归一化(xmin, ymin, xmax, ymax).
https://zh.d2l.ai/chapter_computer-vision/anchor.html
Args:
bb: 真实边界框(bounding box), shape:(nb, 4)
anchor: 待分配的anchor, shape:(na, 4)
jaccard_threshold: 预先设定的阈值
Returns:
assigned_idx: shape: (na, ), 每个anchor分配的真实bb对应的索引, 若未分配任何bb则为-1
"""
na = anchor.shape[0]
nb = bb.shape[0]
jaccard = compute_jaccard(anchor, bb).detach().cpu().numpy() # shape: (na, nb)
assigned_idx = np.ones(na) * -1 # 存放标签初始全为-1
# 先为每个bb分配一个anchor(不要求满足jaccard_threshold)
jaccard_cp = jaccard.copy()
for j in range(nb):
i = np.argmax(jaccard_cp[:, j])
assigned_idx[i] = j
jaccard_cp[i, :] = float("-inf") # 赋值为负无穷, 相当于去掉这一行
# 处理还未被分配的anchor, 要求满足jaccard_threshold
for i in range(na):
if assigned_idx[i] == -1:
j = np.argmax(jaccard[i, :])
if jaccard[i, j] >= jaccard_threshold:
assigned_idx[i] = j
return torch.tensor(assigned_idx, dtype=torch.long)
def xy_to_cxcy(xy):
"""
将(x_min, y_min, x_max, y_max)形式的anchor转换成(center_x, center_y, w, h)形式的.
https://github.com/sgrvinod/a-PyTorch-Tutorial-to-Object-Detection/blob/master/utils.py
Args:
xy: bounding boxes in boundary coordinates, a tensor of size (n_boxes, 4)
Returns:
bounding boxes in center-size coordinates, a tensor of size (n_boxes, 4)
"""
return torch.cat([(xy[:, 2:] + xy[:, :2]) / 2, # c_x, c_y
xy[:, 2:] - xy[:, :2]], 1) # w, h
def MultiBoxTarget(anchor, label):
"""
# 按照「9.4.1. 生成多个锚框」所讲的实现, anchor表示成归一化(xmin, ymin, xmax, ymax).
https://zh.d2l.ai/chapter_computer-vision/anchor.html
Args:
anchor: torch tensor, 输入的锚框, 一般是通过MultiBoxPrior生成, shape:(1,锚框总数,4)
label: 真实标签, shape为(bn, 每张图片最多的真实锚框数, 5)
第二维中,如果给定图片没有这么多锚框, 可以先用-1填充空白, 最后一维中的元素为[类别标签, 四个坐标值]
Returns:
列表, [bbox_offset, bbox_mask, cls_labels]
bbox_offset: 每个锚框的标注偏移量,形状为(bn,锚框总数*4)
bbox_mask: 形状同bbox_offset, 每个锚框的掩码, 一一对应上面的偏移量, 负类锚框(背景)对应的掩码均为0, 正类锚框的掩码均为1
cls_labels: 每个锚框的标注类别, 其中0表示为背景, 形状为(bn,锚框总数)
"""
assert len(anchor.shape) == 3 and len(label.shape) == 3
bn = label.shape[0]
def MultiBoxTarget_one(anc, lab, eps=1e-6):
"""
MultiBoxTarget函数的辅助函数, 处理batch中的一个
Args:
anc: shape of (锚框总数, 4)
lab: shape of (真实锚框数, 5), 5代表[类别标签, 四个坐标值]
eps: 一个极小值, 防止log0
Returns:
offset: (锚框总数*4, )
bbox_mask: (锚框总数*4, ), 0代表背景, 1代表非背景
cls_labels: (锚框总数, 4), 0代表背景
"""
an = anc.shape[0]
# 变量的意义
assigned_idx = assign_anchor(lab[:, 1:], anc) # (锚框总数, )
print("a: ", assigned_idx.shape)
print(assigned_idx)
bbox_mask = ((assigned_idx >= 0).float().unsqueeze(-1)).repeat(1, 4) # (锚框总数, 4)
print("b: " , bbox_mask.shape)
print(bbox_mask)
cls_labels = torch.zeros(an, dtype=torch.long) # 0表示背景
assigned_bb = torch.zeros((an, 4), dtype=torch.float32) # 所有anchor对应的bb坐标
for i in range(an):
bb_idx = assigned_idx[i]
if bb_idx >= 0: # 即非背景
cls_labels[i] = lab[bb_idx, 0].long().item() + 1 # 注意要加一
assigned_bb[i, :] = lab[bb_idx, 1:]
# 如何计算偏移量
center_anc = xy_to_cxcy(anc) # (center_x, center_y, w, h)
center_assigned_bb = xy_to_cxcy(assigned_bb)
offset_xy = 10.0 * (center_assigned_bb[:, :2] - center_anc[:, :2]) / center_anc[:, 2:]
offset_wh = 5.0 * torch.log(eps + center_assigned_bb[:, 2:] / center_anc[:, 2:])
offset = torch.cat([offset_xy, offset_wh], dim = 1) * bbox_mask # (锚框总数, 4)
return offset.view(-1), bbox_mask.view(-1), cls_labels
# 组合输出
batch_offset = []
batch_mask = []
batch_cls_labels = []
for b in range(bn):
offset, bbox_mask, cls_labels = MultiBoxTarget_one(anchor[0, :, :], label[b, :, :])
batch_offset.append(offset)
batch_mask.append(bbox_mask)
batch_cls_labels.append(cls_labels)
bbox_offset = torch.stack(batch_offset)
bbox_mask = torch.stack(batch_mask)
cls_labels = torch.stack(batch_cls_labels)
return [bbox_offset, bbox_mask, cls_labels]
输入预测边界框
在模型预测阶段,我们先为图像生成多个锚框,并为这些锚框一一预测类别和偏移量。随后,我们根据锚框及其预测偏移量得到预测边界框。当锚框数量较多时,同一个目标上可能会输出较多相似的预测边界框。为了使结果更加简洁,我们可以移除相似的预测边界框。常用的方法叫作非极大值抑制(non-maximum suppression,NMS)
非极大值抑制
图像风格迁移
样式迁移
使用卷积神经网络自动将某图像中的样式应用在另一图像之上,即样式迁移(style transfer)。这里我们需要两张输入图像,一张是内容图像,另一张是样式图像,我们将使用神经网络修改内容图像使其在样式上接近样式图像
方法
读取内容图像和样式图像
预处理和后处理图像
rgb_mean = np.array([0.485, 0.456, 0.406])
rgb_std = np.array([0.229, 0.224, 0.225])
def preprocess(PIL_img, image_shape):
process = torchvision.transforms.Compose([
torchvision.transforms.Resize(image_shape),
torchvision.transforms.ToTensor(),
torchvision.transforms.Normalize(mean=rgb_mean, std=rgb_std)])
return process(PIL_img).unsqueeze(dim = 0) # (batch_size, 3, H, W)
def postprocess(img_tensor):
inv_normalize = torchvision.transforms.Normalize(
mean= -rgb_mean / rgb_std,
std= 1/rgb_std)
to_PIL_image = torchvision.transforms.ToPILImage()
return to_PIL_image(inv_normalize(img_tensor[0].cpu()).clamp(0, 1))
抽取特征
def extract_features(X, content_layers, style_layers):
contents = []
styles = []
for i in range(len(net)):
X = net[i](X)
if i in style_layers:
styles.append(X)
if i in content_layers:
contents.append(X)
return contents, styles
定义损失函数
损失函数由内容损失、样式损失和总变差损失三部分组成
内容损失
与线性回归中的损失函数类似,内容损失是通过平方误差函数衡量合成图像和内容图像在内容特征上的差异,平方误差函数的两个输入均为extract_features函数计算所得到的内容层的输出
样式损失
样式损失是也一样通过平方误差函数衡量合成图像与样式图像在样式上的差异
总变差损失
当合成图像中存在大量高频噪点,即有特别亮或者特别暗的颗粒像素,一种常用的降噪方法就是总变差降噪(total variation denoising)
图像分类案例1
Kaggle上的图像分类(CIFAR-10)
define Model
class ResidualBlock(nn.Module): # 我们定义网络时一般是继承的torch.nn.Module创建新的子类
def __init__(self, inchannel, outchannel, stride=1):
super(ResidualBlock, self).__init__()
#torch.nn.Sequential是一个Sequential容器,模块将按照构造函数中传递的顺序添加到模块中。
self.left = nn.Sequential(
nn.Conv2d(inchannel, outchannel, kernel_size=3, stride=stride, padding=1, bias=False),
# 添加第一个卷积层,调用了nn里面的Conv2d()
nn.BatchNorm2d(outchannel), # 进行数据的归一化处理
nn.ReLU(inplace=True), # 修正线性单元,是一种人工神经网络中常用的激活函数
nn.Conv2d(outchannel, outchannel, kernel_size=3, stride=1, padding=1, bias=False),
nn.BatchNorm2d(outchannel)
)
self.shortcut = nn.Sequential()
if stride != 1 or inchannel != outchannel:
self.shortcut = nn.Sequential(
nn.Conv2d(inchannel, outchannel, kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(outchannel)
)
# 便于之后的联合,要判断Y = self.left(X)的形状是否与X相同
def forward(self, x): # 将两个模块的特征进行结合,并使用ReLU激活函数得到最终的特征。
out = self.left(x)
out += self.shortcut(x)
out = F.relu(out)
return out
class ResNet(nn.Module):
def __init__(self, ResidualBlock, num_classes=10):
super(ResNet, self).__init__()
self.inchannel = 64
self.conv1 = nn.Sequential( # 用3个3x3的卷积核代替7x7的卷积核,减少模型参数
nn.Conv2d(3, 64, kernel_size=3, stride=1, padding=1, bias=False),
nn.BatchNorm2d(64),
nn.ReLU(),
)
self.layer1 = self.make_layer(ResidualBlock, 64, 2, stride=1)
self.layer2 = self.make_layer(ResidualBlock, 128, 2, stride=2)
self.layer3 = self.make_layer(ResidualBlock, 256, 2, stride=2)
self.layer4 = self.make_layer(ResidualBlock, 512, 2, stride=2)
self.fc = nn.Linear(512, num_classes)
def make_layer(self, block, channels, num_blocks, stride):
strides = [stride] + [1] * (num_blocks - 1) #第一个ResidualBlock的步幅由make_layer的函数参数stride指定
# ,后续的num_blocks-1个ResidualBlock步幅是1
layers = []
for stride in strides:
layers.append(block(self.inchannel, channels, stride))
self.inchannel = channels
return nn.Sequential(*layers)
def forward(self, x):
out = self.conv1(x)
out = self.layer1(out)
out = self.layer2(out)
out = self.layer3(out)
out = self.layer4(out)
out = F.avg_pool2d(out, 4)
out = out.view(out.size(0), -1)
out = self.fc(out)
return out
def ResNet18():
return ResNet(ResidualBlock)
Training and Testing
# 定义是否使用GPU
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
# 超参数设置
EPOCH = 20 #遍历数据集次数
pre_epoch = 0 # 定义已经遍历数据集的次数
LR = 0.1 #学习率
# 模型定义-ResNet
net = ResNet18().to(device)
# 定义损失函数和优化方式
criterion = nn.CrossEntropyLoss() #损失函数为交叉熵,多用于多分类问题
optimizer = optim.SGD(net.parameters(), lr=LR, momentum=0.9, weight_decay=5e-4)
#优化方式为mini-batch momentum-SGD,并采用L2正则化(权重衰减)
# 训练
if __name__ == "__main__":
print("Start Training, Resnet-18!")
num_iters = 0
for epoch in range(pre_epoch, EPOCH):
print('\nEpoch: %d' % (epoch + 1))
net.train()
sum_loss = 0.0
correct = 0.0
total = 0
for i, data in enumerate(trainloader, 0):
#用于将一个可遍历的数据对象(如列表、元组或字符串)组合为一个索引序列,同时列出数据和数据下标,
#下标起始位置为0,返回 enumerate(枚举) 对象。
num_iters += 1
inputs, labels = data
inputs, labels = inputs.to(device), labels.to(device)
optimizer.zero_grad() # 清空梯度
# forward + backward
outputs = net(inputs)
loss = criterion(outputs, labels)
loss.backward()
optimizer.step()
sum_loss += loss.item() * labels.size(0)
_, predicted = torch.max(outputs, 1) #选出每一列中最大的值作为预测结果
total += labels.size(0)
correct += (predicted == labels).sum().item()
# 每20个batch打印一次loss和准确率
if (i + 1) % 20 == 0:
print('[epoch:%d, iter:%d] Loss: %.03f | Acc: %.3f%% '
% (epoch + 1, num_iters, sum_loss / (i + 1), 100. * correct / total))
print("Training Finished, TotalEPOCH=%d" % EPOCH)