chaehoyu/clock-vae-color-140x-v1

⏰ Clock-VAE-Color-140x

특정 시간의 아날로그 시계 이미지를 생성하기 위해 설계된 Conditional VAE 모델입니다.
A Conditional VAE tailored for generating analog clock images that represent specific times.

---

📛 Naming Convention

• clock-vae: 모델 이름 (Model name)
• color: 이미지 유형 (Image type: color or mono)
• 140x: 이미지 크기 (Image size: 140x140)
• v1: 모델 버전 (Model version)

---

🔧 Model Definition Code

class ConditionalVAE(nn.Module):
    def __init__(self, input_dim, condition_dim, latent_dim):
        super(ClockVAEHandler.ConditionalVAE, self).__init__()
        self.encoder = nn.Sequential(
            nn.Linear(input_dim + condition_dim, 512),
            nn.ReLU(),
            nn.Linear(512, 256),
            nn.ReLU(),
            nn.Linear(256, 128),
            nn.ReLU(),
        )
        self.fc_mu = nn.Linear(128, latent_dim)
        self.fc_logvar = nn.Linear(128, latent_dim)
        self.decoder = nn.Sequential(
            nn.Linear(latent_dim + condition_dim, 128),
            nn.ReLU(),
            nn.Linear(128, 256),
            nn.ReLU(),
            nn.Linear(256, 512),
            nn.ReLU(),
            nn.Linear(512, input_dim),
            nn.Sigmoid()
        )

    def encode(self, x, condition):
        x = x.view(x.size(0), -1)
        condition = condition.view(condition.size(0), -1)
        x_cond = torch.cat([x, condition], dim=1)
        h = self.encoder(x_cond)
        mu = self.fc_mu(h)
        logvar = self.fc_logvar(h)
        return mu, logvar

    def reparameterize(self, mu, logvar):
        std = torch.exp(0.5 * logvar)
        eps = torch.randn_like(std)
        return mu + eps * std

    def decode(self, z, condition):
        z_cond = torch.cat([z, condition], dim=1)
        return self.decoder(z_cond)

    def forward(self, x, condition):
        mu, logvar = self.encode(x, condition)
        z = self.reparameterize(mu, logvar)
        return self.decode(z, condition), mu, logvar