arquiteturaREVISAR.py

import torch
import torch.nn as nn
import torch.nn.functional as F

class ResidualBlock(nn.Module):
    def __init__(self, in_channels, out_channels):
        super(ResidualBlock, self).__init__()
        self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, padding=1)
        self.bn1 = nn.BatchNorm2d(out_channels)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, padding=1)
        self.bn2 = nn.BatchNorm2d(out_channels)

    def forward(self, x):
        residual = x
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)
        out = self.conv2(out)
        out = self.bn2(out)
        out += residual
        out = self.relu(out)
        return out

class LicensePlateRecognitionNet(nn.Module):
    def __init__(self):
        super(LicensePlateRecognitionNet, self).__init__()

        self.initial_block = nn.Sequential(
            nn.Conv2d(3, 64, kernel_size=3, padding=1),
            nn.BatchNorm2d(64),
            nn.ReLU(inplace=True),
            nn.Conv2d(64, 128, kernel_size=3, padding=1),
            nn.BatchNorm2d(128),
            nn.ReLU(inplace=True),
        )

        self.residual_blocks = nn.Sequential(
            *[ResidualBlock(128, 128) for _ in range(14)]
        )

        self.conv_block = nn.Sequential(
            nn.MaxPool2d(kernel_size=2, stride=2),
            nn.Conv2d(128, 256, kernel_size=3, padding=1),
            nn.BatchNorm2d(256),
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=2, stride=2),
            nn.Conv2d(256, 512, kernel_size=3, padding=1),
            nn.BatchNorm2d(512),
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=2, stride=2),
            nn.Conv2d(512, 1024, kernel_size=3, padding=1),
            nn.BatchNorm2d(1024),
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=2, stride=2),
        )

        self.detection = nn.Conv2d(1024, 8, kernel_size=1)  # 2 for object/non-object probabilities + 6 for affine transformation parameters

    def forward(self, x):
        out = self.initial_block(x)
        out = self.residual_blocks(out)
        out = self.conv_block(out)
        out = self.detection(out)
        return out


model = LicensePlateRecognitionNet()
input_image = torch.randn(1, 3, 256, 256)  
output = model(input_image)

print("Output shape:", output.shape)