李宏毅GAN学习笔记（02）

GAN Lecture 2

---

Conditional Generation by GAN

Algorithm

In each traing iteration:

• Sample m positive examples $\{(c^1, x^1), (c^2, x^2), \dots, (c^m, x^m)\}${(c1,x1),(c2,x2),…,(cm,xm)} from database
• Sample m noise samples $\{z^1, z^2, \dots, z^m\}${z1,z2,…,zm} from a distribution
• Obtaining generated data $\{\tilde{x}^1, \tilde{x}^2, \dots, \tilde{x}^m\}${x~1,x~2,…,x~m}, $\tilde{x}^i=G(c^i, z^i)$x~i=G(ci,zi)
• Sample m objects $\{\hat{x}^1, \hat{x}^2, \dots, \hat{x}^m\}${x^1,x^2,…,x^m} from database
• Update discriminator parameters $\theta_d$θd​ to maximize
  • $\tilde{V}=\frac{1}{m}\sum_{i=1}^mlogD(c^i, x^i)+\frac{1}{m}\sum_{i=1}^mlog(1-D(c^i, \tilde{x}^i))+\frac{1}{m}_{i=1}^mlog(1-D(c^i, \hat{x}^i))$V~=m1​∑i=1m​logD(ci,xi)+m1​∑i=1m​log(1−D(ci,x~i))+m1​i=1m​log(1−D(ci,x^i))
  • $\theta_d \leftarrow \theta_d+\eta\bigtriangledown\tilde{V}(\theta_d)$θd​←θd​+η▽V~(θd​)

Learning D

• Sample m noise samples $\{z^1,z^2,\dots,z^m\}${z1,z2,…,zm} from a distribution
• Sample m conditions $\{c^1,c^2,\dots,c^m\}${c1,c2,…,cm} from a database
• Update generator parameters $\theta_g$θg​ to maximize
  • $\tilde{V}=\frac{1}{m}\sum_{i=1}^mlog(D(G(c^i, z^i)))$V~=m1​∑i=1m​log(D(G(ci,zi))), $\theta_g \leftarrow\eta\bigtriangledown\tilde{V}(\theta_g)$θg​←η▽V~(θg​)

Learning G

倾向推荐第二种网络架构
参考文献：StackGAN

参考文献：Patch GAN

参考例子：Github