GAN_Metrics-Tensorflow
Simple Tensorflow implementation of metrics for GAN evaluation
- Inception score
- Frechet-Inception distance
- Kernel-Inception distance
Summary
Name | Description | Performance score |
---|---|---|
Inception score | KL-Divergence between conditional and marginal label distributions over generated data. | Higher is better. |
Frechet-Inception distance | Wasserstein-2 distance between multi-variate Gaussians fitted to data embedded into a feature space. | Lower is better. |
Kernel-Inception distance | Measures the dissimilarity between two probability distributions Pr and Pg using samples drawn independently from each distribution. |
Lower is better. |
Usage
βββ real_source
βββ aaa.png
βββ bbb.jpg
βββ real_target
βββ ccc.png
βββ ddd.jpg
βββ fake
βββ ccc_fake.png
βββ ddd_fake.jpg
βββ main.py
βββ inception_score.py
βββ frechet_kernel_Inception_distance.py
> python main.py
Reference
Author
Junho Kim