Generative Adversarial Networks (GANs) are powerful generative models thatachieved strong results, mainly in the image domain . Evolutionary algorithms, such as COEGAN, were recently proposed as a solution to improve the GAN training . We propose an evaluation method based on t-distributed Stochastic NeighbourEmbedding (t-SNE) to assess the progress of GANs and visualize the distributionlearned by generators in training . A metric based on the resulting t-SNEmaps and the Jaccard index is proposed to represent the model quality . The results show both by visual inspection and metrics that theEvolutionary Algorithm gradually improves discriminators and generators throughgenerations, avoiding problems such as mode collapse, avoid problems such .

Author(s) : Victor Costa, Nuno Lourenço, João Correia, Penousal Machado

Links : PDF - Abstract

Code :

https://github.com/syahdeini/gan


Coursera

Keywords : gans - results - based - generators - generative -

Leave a Reply

Your email address will not be published. Required fields are marked *