Generative Adversarial Networks
GANs (mainly in image synthesis)
- Assumption
- MEANINGFUL: The generated image should be clear, the output probability of a classifier network should be [0.9, 0.05, ...] (largely skewed to a class).is of low entropy.
- DIVERSITY: If we have 10 classes, the generated image should be averagely distributed. So that the marginal distribution__is of high entropy.
- Better models: KL Divergence ofandshould be high.
- Formulation
-
- where
- is sampled from generated data
- is the output probability of Inception v3 when input is
- is the average output probability of all generated data (from InceptionV3, 1000-dim vector)
- , whereis the dimension of the output probability.
- Reference
- Formulation
-
- where
- andare the 2048-dim activations the Inception v3 pool3 layer
- is the mean of real photo's feature
- is the mean of generated photo's feature
- is the covariance matrix of real photo's feature
- is the covariance matrix of generated photo's feature
- Reference