질문 및 토론

gan_title.png

🔗 논문 링크

Generative Adversarial Networks

👨‍💻 구현 코드

gan.ipynb

📚 참고한 자료

1시간만에 GAN(Generative Adversarial Network) 완전 정복하기

[목차]

Abstract

Adversarial nets

$$ \min_G \max_D V(D, G) = \mathbb{E}{\mathbf{x} \sim p{data}(\mathbf{x})} [\log D(\mathbf{x})] + \mathbb{E}{\mathbf{z} \sim p{\mathbf{z}}(\mathbf{z})} [\log (1 - D(G(\mathbf{z})))] $$

algorithm_1.png

Theoretical Results

$$ \min_G \max_D V(D, G) = \mathbb{E}{\mathbf{x} \sim p{data}(\mathbf{x})} [\log D(\mathbf{x})] + \mathbb{E}{\mathbf{z} \sim p{\mathbf{z}}(\mathbf{z})} [\log (1 - D(G(\mathbf{z})))] \tag{1} $$

$$ \begin{align} \min_{G} V(D^*G, G) &= -\log(4) + KL \left( p{data} \| \dfrac{p_{data} + p_g}{2} \right) + KL \left( p_{g} \| \dfrac{p_{data} + p_g}{2} \right) \\ &= -\log(4) + 2 \cdot JSD(p_{data} \| p_g) \tag{2} \end{align} $$