|Liqun Chen||Duke University|
|Shuyang Dai||Duke University|
|Chenyang Tao||Duke University|
|Haichao Zhang||Baidu Research|
|Dinghan Shen||Duke University|
|Yizhe Zhang||Microsoft Research|
|Guoyin Wang||Duke University|
|Ruiyi Zhang||Duke University|
|Lawrence Carin||Duke University|
Generative adversarial networks (GANs) have achieved significant success in generating real-valued data.Instead of using the standard GAN objective, the authors propose to improve text-generation GAN via a novel approach inspired by optimal transport.
Generative adversarial networks (GANs) have achieved significant success in generating real-valued data. However, the discrete nature of text hinders the application of GAN to text-generation tasks. Instead of using the standard GAN objective, we propose to improve text-generation GAN via a novel approach inspired by optimal transport. Specifically, we consider matching the latent feature distributions of real and synthetic sentences using a novel metric, termed the feature-mover's distance (FMD). This formulation leads to a highly discriminative critic and easy-to-optimize objective, overcoming the mode-collapsing and brittle-training problems in existing methods. Extensive experiments are conducted on a variety of tasks to evaluate the proposed model empirically, including unconditional text generation, style transfer from non-parallel text, and unsupervised cipher cracking. The proposed model yields superior performance, demonstrating wide applicability and effectiveness.