fast rcnn代码

以下是 Fast R-CNN 的 PyTorch 代码示例：

import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision.models as models

class FastRCNN(nn.Module):
    def __init__(self, num_classes):
        super(FastRCNN, self).__init__()
        self.features = models.vgg16(pretrained=True).features
        self.roi_pool = nn.AdaptiveMaxPool2d((7, 7))
        self.fc1 = nn.Linear(512 * 7 * 7, 4096)
        self.fc2 = nn.Linear(4096, 4096)
        self.cls_score = nn.Linear(4096, num_classes)
        self.bbox_pred = nn.Linear(4096, num_classes * 4)

    def forward(self, x, rois):
        x = self.features(x)
        rois = torch.cat([rois[:, :1] * x.size(3), rois[:, 1:2] * x.size(2), rois[:, 2:3] * x.size(3), rois[:, 3:] * x.size(2)], 1)
        crops = torch.zeros(rois.size(0), x.size(1), 7, 7).to(x.device)
        for i in range(rois.size(0)):
            crops[i] = self.roi_pool(x[:, :, int(rois[i, 1]):int(rois[i, 3]) + 1, int(rois[i, 0]):int(rois[i, 2]) + 1])
        x = crops.view(crops.size(0), -1)
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        cls_score = self.cls_score(x)
        bbox_pred = self.bbox_pred(x)
        return cls_score, bbox_pred

这个代码实现了一个 Fast R-CNN 模型，并使用 VGG16 作为特征提取器。在前向传播过程中，给定输入图像和 RoI，模型首先通过特征提取器提取特征图，然后通过 RoI 池化层从特征图中提取每个 RoI 的特征，将其展平并通过全连接层进行分类和边界框回归。