Pytorch深度学习【十六】

物体检测

边缘框

• 一个边缘框可以通过4个数字定义
  • (左上x,左上y,右下x,右下y)
  • (左上x,左上y,宽,高)
• 目标检测数据集
  • 每行表示一个物体
    • 图片文件名，物体类别，边缘框
  • COCO
    • 80物体，330K图片，1.5M物体
• 总结
  • 物体检测识别图片里的多个物体的类别和位置
  • 位置通常用边缘框表示
• 边缘框相关操作的代码实现

%matplotlib inline
import torch
from d2l import torch as d2l

d2l.set_figsize()
img = d2l.plt.imread('../img/catdog.jpg')
d2l.plt.imshow(img);
# 两种变换的表示
def box_corner_to_center(boxes):
    """从（左上，右下）转换到（中间，宽度，高度）"""
    x1, y1, x2, y2 = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3]
    cx = (x1 + x2) / 2
    cy = (y1 + y2) / 2
    w = x2 - x1
    h = y2 - y1
    boxes = torch.stack((cx, cy, w, h), axis=-1)
    return boxes

def box_center_to_corner(boxes):
    """从（中间，宽度，高度）转换到（左上，右下）"""
    cx, cy, w, h = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3]
    x1 = cx - 0.5 * w
    y1 = cy - 0.5 * h
    x2 = cx + 0.5 * w
    y2 = cy + 0.5 * h
    boxes = torch.stack((x1, y1, x2, y2), axis=-1)
    return boxes
# 给定边框测试
dog_bbox, cat_bbox = [60.0, 45.0, 378.0, 516.0], [400.0, 112.0, 655.0, 493.0]
boxes = torch.tensor((dog_bbox, cat_bbox))
box_center_to_corner(box_corner_to_center(boxes)) == boxes
# 将边框在具体位置处画出来
def bbox_to_rect(bbox, color):
    return d2l.plt.Rectangle(xy=(bbox[0], bbox[1]), width=bbox[2] - bbox[0],
                             height=bbox[3] - bbox[1], fill=False,
                             edgecolor=color, linewidth=2)
fig = d2l.plt.imshow(img)
fig.axes.add_patch(bbox_to_rect(dog_bbox, 'blue'))
fig.axes.add_patch(bbox_to_rect(cat_bbox, 'red'));

• 对目标检测数据集的操作

# 读取香蕉数据集
def read_data_bananas(is_train=True):
    """读取香蕉检测数据集中的图像和标签。"""
    data_dir = d2l.download_extract('banana-detection')
    csv_fname = os.path.join(data_dir,
                             'bananas_train' if is_train else 'bananas_val',
                             'label.csv')
    csv_data = pd.read_csv(csv_fname)
    csv_data = csv_data.set_index('img_name')
    images, targets = [], []
    for img_name, target in csv_data.iterrows():
        images.append(
            torchvision.io.read_image(
                os.path.join(data_dir,
                             'bananas_train' if is_train else 'bananas_val',
                             'images', f'{img_name}')))
        targets.append(list(target))
    return images, torch.tensor(targets).unsqueeze(1) / 256
# 创建一个自定义 Dataset 实例
class BananasDataset(torch.utils.data.Dataset):
    """一个用于加载香蕉检测数据集的自定义数据集。"""
    def __init__(self, is_train):
        self.features, self.labels = read_data_bananas(is_train)
        print('read ' + str(len(self.features)) + (
            f' training examples' if is_train else f' validation examples'))
# 迭代后被自动调用
    def __getitem__(self, idx):
        return (self.features[idx].float(), self.labels[idx])

    def __len__(self):
        return len(self.features)
# 为训练集和测试集返回两个数据加载器实例
def load_data_bananas(batch_size):
    """加载香蕉检测数据集。"""
    train_iter = torch.utils.data.DataLoader(BananasDataset(is_train=True),
                                             batch_size, shuffle=True)
    val_iter = torch.utils.data.DataLoader(BananasDataset(is_train=False),
                                           batch_size)
    return train_iter, val_iter
# 读取一个小批量，并打印其中的图像和标签的形状
batch_size, edge_size = 32, 256
train_iter, _ = load_data_bananas(batch_size)
batch = next(iter(train_iter))
batch[0].shape, batch[1].shape
# 示例
imgs = (batch[0][0:10].permute(0, 2, 3, 1)) / 255
axes = d2l.show_images(imgs, 2, 5, scale=2)
for ax, label in zip(axes, batch[1][0:10]):
    d2l.show_bboxes(ax, [label[0][1:5] * edge_size], colors=['w'])

锚框

• 定义
  • 一类目标检测算法是基于锚框
    • 提出多个被称为锚框的区域（边缘框）
    • 预测每个锚框里是否含有关注的物体
    • 如果是，预测从这个锚框到真实边缘框的偏移
• IoU-交并比—从锚框拟合到实际预测框
  • 本意用来计算两个框的相似度
    $$J(A,B)=\frac{|A\bigcap B|}{｜A\bigcup B｜}$$
• 赋予锚框标号
  • 每个锚框是一个训练样本
  • 将每个锚框，要么标准成背景，关联上一个真实边缘框
  • 我们可能会生成大量的锚框
    • 这个导致大量的负类样本出现
• 通俗理解锚框
  • 锚框就是我们希望拟合到已经标注的边缘框的训练样本
  • 锚框本身就是一种训练样本，因此，一张图片可以定义多个锚框，使得其训练多次
• 使用非极大值抑制(NMS)输出
  • 每个锚框预测一个边缘框
  • NMS可以合并相似的预测
    • 选中是非背景锚框，其为最大预测值
    • 去掉所有其他和它IOU值大于$\theta$的预测
    • 重复上述过程直到所有预测要么被选中，要么被去掉
• 总结
  • 锚框是自定义的训练样本，边缘框是已经标注好的样本
  • 全流程分析
    • 锚框与边缘框求IOU，保证至少一个边缘框对应一个锚框，其他IOU不满足域值的设为背景框
    • 被选出的锚框中，将每一类物体最大值的锚框选出，并且与同预测类的进行IOU求解，IOU大于某一域值的全部被滤除，这样在一个目标周围就保证了只出现一个框
  • 官方思路
    • 首先生成大量锚框，并赋予编号，每个锚框进行一个样本进行训练
    • 在预测时，使用NMS来去掉冗余的预测
• 代码锚框相关技术

%matplotlib inline
import torch
from d2l import torch as d2l

torch.set_printoptions(2) # 设置浮点数输出几位
def multibox_prior(data, sizes, ratios):
    """生成以每个像素为中心具有不同形状的锚框。"""
    in_height, in_width = data.shape[-2:]
    device, num_sizes, num_ratios = data.device, len(sizes), len(ratios)
    boxes_per_pixel = (num_sizes + num_ratios - 1) # 每一个像素点生成多少个框
    size_tensor = torch.tensor(sizes, device=device)
    ratio_tensor = torch.tensor(ratios, device=device)

    offset_h, offset_w = 0.5, 0.5
    steps_h = 1.0 / in_height
    steps_w = 1.0 / in_width

    center_h = (torch.arange(in_height, device=device) + offset_h) * steps_h
    center_w = (torch.arange(in_width, device=device) + offset_w) * steps_w
    shift_y, shift_x = torch.meshgrid(center_h, center_w)
    shift_y, shift_x = shift_y.reshape(-1), shift_x.reshape(-1)

    w = torch.cat((size_tensor * torch.sqrt(ratio_tensor[0]),
                   sizes[0] * torch.sqrt(ratio_tensor[1:])))\
                   * in_height / in_width
    h = torch.cat((size_tensor / torch.sqrt(ratio_tensor[0]),
                   sizes[0] / torch.sqrt(ratio_tensor[1:])))
    anchor_manipulations = torch.stack(
        (-w, -h, w, h)).T.repeat(in_height * in_width, 1) / 2

    out_grid = torch.stack([shift_x, shift_y, shift_x, shift_y],
                           dim=1).repeat_interleave(boxes_per_pixel, dim=0)
    output = out_grid + anchor_manipulations
    return output.unsqueeze(0) # 在第0维处加入新的维度

# 返回的锚框变量Y的形状
img = d2l.plt.imread('../img/catdog.jpg')
h, w = img.shape[:2]
print(h, w)
X = torch.rand(size=(1, 3, h, w))
Y = multibox_prior(X, sizes=[0.75, 0.5, 0.25], ratios=[1, 2, 0.5])
# size就是锚框的大小占图片的多少，ratios就是高宽比
Y.shape
# 访问以 (250, 250) 为中心的第一个锚框
boxes = Y.reshape(h, w, 5, 4)
boxes[250, 250, 0, :]
# 0就是第一个锚框，：就是读取数据左上右下的坐标
# 显示以图像中一个像素为中心的所有锚框
def show_bboxes(axes, bboxes, labels=None, colors=None):# 画出所给boxes的框框
    """显示所有边界框。"""
    def _make_list(obj, default_values=None):
        if obj is None:
            obj = default_values
        elif not isinstance(obj, (list, tuple)):
            obj = [obj]
        return obj

    labels = _make_list(labels)
    colors = _make_list(colors, ['b', 'g', 'r', 'm', 'c'])
    for i, bbox in enumerate(bboxes):
        color = colors[i % len(colors)]
        rect = d2l.bbox_to_rect(bbox.detach().numpy(), color)
        axes.add_patch(rect)
        if labels and len(labels) > i:
            text_color = 'k' if color == 'w' else 'w'
            axes.text(rect.xy[0], rect.xy[1], labels[i], va='center',
                      ha='center', fontsize=9, color=text_color,
                      bbox=dict(facecolor=color, lw=0))
d2l.set_figsize()
bbox_scale = torch.tensor((w, h, w, h))
fig = d2l.plt.imshow(img)
show_bboxes(fig.axes, boxes[250, 250, :, :] * bbox_scale, [
    's=0.75, r=1', 's=0.5, r=1', 's=0.25, r=1', 's=0.75, r=2', 's=0.75, r=0.5'
])# 250，250像素点所有锚框都画出来
# 交并比(IoU)
def box_iou(boxes1, boxes2):
    """计算两个锚框或边界框列表中成对的交并比。"""
    box_area = lambda boxes: ((boxes[:, 2] - boxes[:, 0]) *
                              (boxes[:, 3] - boxes[:, 1]))
    areas1 = box_area(boxes1)
    areas2 = box_area(boxes2)
    inter_upperlefts = torch.max(boxes1[:, None, :2], boxes2[:, :2])
    inter_lowerrights = torch.min(boxes1[:, None, 2:], boxes2[:, 2:])
    inters = (inter_lowerrights - inter_upperlefts).clamp(min=0)
    inter_areas = inters[:, :, 0] * inters[:, :, 1]
    union_areas = areas1[:, None] + areas2 - inter_areas
    return inter_areas / union_areas
# 将真实边界框分配给锚框
def assign_anchor_to_bbox(ground_truth, anchors, device, iou_threshold=0.5):
    """将最接近的真实边界框分配给锚框。"""
    num_anchors, num_gt_boxes = anchors.shape[0], ground_truth.shape[0]
    jaccard = box_iou(anchors, ground_truth) # 所有边框和真实边框全部算出来
    anchors_bbox_map = torch.full((num_anchors,), -1, dtype=torch.long,
                                  device=device)
    max_ious, indices = torch.max(jaccard, dim=1)
    anc_i = torch.nonzero(max_ious >= 0.5).reshape(-1)
    box_j = indices[max_ious >= 0.5]
    anchors_bbox_map[anc_i] = box_j
    col_discard = torch.full((num_anchors,), -1)
    row_discard = torch.full((num_gt_boxes,), -1)
    for _ in range(num_gt_boxes):
        max_idx = torch.argmax(jaccard) # 每次把最大的找出来
        box_idx = (max_idx % num_gt_boxes).long()
        anc_idx = (max_idx / num_gt_boxes).long()
        anchors_bbox_map[anc_idx] = box_idx
        jaccard[:, box_idx] = col_discard
        jaccard[anc_idx, :] = row_discard # 删除行列
    return anchors_bbox_map
# 利用公式对偏移量进行定义---标记类和边缘
def offset_boxes(anchors, assigned_bb, eps=1e-6):
    """对锚框偏移量的转换。"""
    c_anc = d2l.box_corner_to_center(anchors)
    c_assigned_bb = d2l.box_corner_to_center(assigned_bb)
    offset_xy = 10 * (c_assigned_bb[:, :2] - c_anc[:, :2]) / c_anc[:, 2:]
    offset_wh = 5 * torch.log(eps + c_assigned_bb[:, 2:] / c_anc[:, 2:])
    offset = torch.cat([offset_xy, offset_wh], axis=1)
    return offset
# 标记锚框的类和偏移量
def multibox_target(anchors, labels):
    """使用真实边界框标记锚框。"""
    batch_size, anchors = labels.shape[0], anchors.squeeze(0)
    batch_offset, batch_mask, batch_class_labels = [], [], []
    device, num_anchors = anchors.device, anchors.shape[0]
    for i in range(batch_size):
        label = labels[i, :, :]
        anchors_bbox_map = assign_anchor_to_bbox(label[:, 1:], anchors,
                                                 device)
        bbox_mask = ((anchors_bbox_map >= 0).float().unsqueeze(-1)).repeat(
            1, 4)
        class_labels = torch.zeros(num_anchors, dtype=torch.long,
                                   device=device)
        assigned_bb = torch.zeros((num_anchors, 4), dtype=torch.float32,
                                  device=device)
        indices_true = torch.nonzero(anchors_bbox_map >= 0)
        bb_idx = anchors_bbox_map[indices_true]
        class_labels[indices_true] = label[bb_idx, 0].long() + 1
        assigned_bb[indices_true] = label[bb_idx, 1:]
        offset = offset_boxes(anchors, assigned_bb) * bbox_mask
        batch_offset.append(offset.reshape(-1))
        batch_mask.append(bbox_mask.reshape(-1))
        batch_class_labels.append(class_labels)
    bbox_offset = torch.stack(batch_offset)
    bbox_mask = torch.stack(batch_mask)
    class_labels = torch.stack(batch_class_labels)
    return (bbox_offset, bbox_mask, class_labels)
    # bbox_offset反应真实框和预测锚框的差距，bbox_mask为0则认为其本身就是背景，class_labels是锚框具体的类
# 在图像中绘制这些地面真相边界框和锚框
ground_truth = torch.tensor([[0, 0.1, 0.08, 0.52, 0.92],
                             [1, 0.55, 0.2, 0.9, 0.88]])
anchors = torch.tensor([[0, 0.1, 0.2, 0.3], [0.15, 0.2, 0.4, 0.4],
                        [0.63, 0.05, 0.88, 0.98], [0.66, 0.45, 0.8, 0.8],
                        [0.57, 0.3, 0.92, 0.9]])

fig = d2l.plt.imshow(img)
show_bboxes(fig.axes, ground_truth[:, 1:] * bbox_scale, ['dog', 'cat'], 'k')
show_bboxes(fig.axes, anchors * bbox_scale, ['0', '1', '2', '3', '4']);
# 根据狗和猫的真实边界框，标注这些锚框的分类和偏移量
labels = multibox_target(anchors.unsqueeze(dim=0),
                         ground_truth.unsqueeze(dim=0))
labels[2] # 输出每个框最后表达的类
labels[1] # 是否为背景
labels[0] # 与真实框的差距
# 应用逆偏移变换来返回预测的边界框坐标
def offset_inverse(anchors, offset_preds):
    """根据带有预测偏移量的锚框来预测边界框。"""
    anc = d2l.box_corner_to_center(anchors)
    pred_bbox_xy = (offset_preds[:, :2] * anc[:, 2:] / 10) + anc[:, :2]
    pred_bbox_wh = torch.exp(offset_preds[:, 2:] / 5) * anc[:, 2:]
    pred_bbox = torch.cat((pred_bbox_xy, pred_bbox_wh), axis=1)
    predicted_bbox = d2l.box_center_to_corner(pred_bbox)
    return predicted_bbox
# 以下 nms 函数按降序对置信度进行排序并返回其索引
def nms(boxes, scores, iou_threshold):
    """对预测边界框的置信度进行排序。"""
    B = torch.argsort(scores, dim=-1, descending=True)
    keep = []
    while B.numel() > 0:
        i = B[0]
        keep.append(i)
        if B.numel() == 1: break
        iou = box_iou(boxes[i, :].reshape(-1, 4),
                      boxes[B[1:], :].reshape(-1, 4)).reshape(-1)
        inds = torch.nonzero(iou <= iou_threshold).reshape(-1)
        B = B[inds + 1]
    return torch.tensor(keep, device=boxes.device)
# 将非极大值抑制应用于预测边界框
def multibox_detection(cls_probs, offset_preds, anchors, nms_threshold=0.5,
                       pos_threshold=0.009999999):
    """使用非极大值抑制来预测边界框。"""
    device, batch_size = cls_probs.device, cls_probs.shape[0]
    anchors = anchors.squeeze(0)
    num_classes, num_anchors = cls_probs.shape[1], cls_probs.shape[2]
    out = []
    for i in range(batch_size):
        cls_prob, offset_pred = cls_probs[i], offset_preds[i].reshape(-1, 4)
        conf, class_id = torch.max(cls_prob[1:], 0)
        predicted_bb = offset_inverse(anchors, offset_pred)
        keep = nms(predicted_bb, conf, nms_threshold)

        all_idx = torch.arange(num_anchors, dtype=torch.long, device=device)
        combined = torch.cat((keep, all_idx))
        uniques, counts = combined.unique(return_counts=True)
        non_keep = uniques[counts == 1]
        all_id_sorted = torch.cat((keep, non_keep))
        class_id[non_keep] = -1
        class_id = class_id[all_id_sorted]
        conf, predicted_bb = conf[all_id_sorted], predicted_bb[all_id_sorted]
        below_min_idx = (conf < pos_threshold)
        class_id[below_min_idx] = -1
        conf[below_min_idx] = 1 - conf[below_min_idx]
        pred_info = torch.cat(
            (class_id.unsqueeze(1), conf.unsqueeze(1), predicted_bb), dim=1)
        out.append(pred_info)
    return torch.stack(out)
# 将上述算法应用到一个带有四个锚框的具体示例中
anchors = torch.tensor([[0.1, 0.08, 0.52, 0.92], [0.08, 0.2, 0.56, 0.95],
                        [0.15, 0.3, 0.62, 0.91], [0.55, 0.2, 0.9, 0.88]])
offset_preds = torch.tensor([0] * anchors.numel())
cls_probs = torch.tensor([[0] * 4,
                          [0.9, 0.8, 0.7, 0.1],
                          [0.1, 0.2, 0.3, 0.9]])
# 在图像上绘制这些预测边界框和置信度
fig = d2l.plt.imshow(img)
show_bboxes(fig.axes, anchors * bbox_scale,
            ['dog=0.9', 'dog=0.8', 'dog=0.7', 'cat=0.9'])
# 返回结果的形状是（批量大小，锚框的数量，6）
output = multibox_detection(cls_probs.unsqueeze(dim=0),
                            offset_preds.unsqueeze(dim=0),
                            anchors.unsqueeze(dim=0), nms_threshold=0.5)
output
# 输出由非极大值抑制保存的最终预测边界框
fig = d2l.plt.imshow(img)
for i in output[0].detach().numpy():
    if i[0] == -1:
        continue
    label = ('dog=', 'cat=')[int(i[0])] + str(i[1])
    show_bboxes(fig.axes, [torch.tensor(i[2:]) * bbox_scale], label)

目标检测的概括

• 区域卷积神经网络
  • R-CNN—对每一个锚框进行CNN抽取
    • 使用启发式搜索算法来选择锚框
    • 使用预训练模型来对每个锚框提取特征
    • 训练一个SVM来对类别分类
    • 训练一个线性回归模型来预测边缘框偏移
  • ROI兴趣区域的池化
    • 给定一个锚框，均匀分割成n*m块，输出每块里的最大值
    • 不管锚框多大，最后总是输出n*m个值
  • Fast-RCNN—对全局进行CNN抽取特征后再进行锚框的ROI特征
    • 使用CNN对图片抽取特征
    • 使用ROI池化层对每个锚框生成固定长度特征
  • Faster-R-CNN
    • 使用一个区域提议网络来替代启发式搜索来获得更好的锚框
    • 流程—CNN提取特征—RPN进行二分类问题判读是否是一个合理的锚框—ROI
  • Mask-R-CNN
    • 如果有像素级别的标号，使用FCN来利用这些信息
• 单发多框检测(SSD)
  • 核心—端到端的预测性
  • 生成锚框
    • 对每个像素，生成多个以它为中心的锚框
    • 给定$n$个大小$s_1,s_2,\dots ,s_n$和m个高宽比，那么生成n+m-1个锚框，其大小和高宽比分别为
      $$(s_1,r_1),(s_2,r_1),\dots ,(s_n,r_1),\dots ,(s_1,r_m)$$
  • 思路
    • 一个基础网络来抽取特征，然后多个卷积层块来减半高宽
    • 在每段都生成锚框
      • 底部段来拟合小物体，顶部段来拟合大物体
    • 对每个锚框预测类别和边缘框
  • 总结
    • SSD通过单神经网络来检测模型
    • 以每个像素为中心的产生多个锚框
    • 在多个段的输出上进行多尺度的检测
• YOLO
  • SSD中锚框大量重叠，因为每个像素进行锚框，因此浪费来很多计算
  • YOLO将图片直接均匀分成了S*S个锚框
  • 每个锚框预测$B$个边缘框
  • 后续版本(v2,v3,v4,$\dots)