Poster
Toward Controlled Generation of Text
Zhiting Hu · Zichao Yang · Xiaodan Liang · Ruslan Salakhutdinov · Eric Xing

Wed Aug 9th 06:30 -- 10:00 PM @ Gallery #80

Generic generation and manipulation of text is challenging and has limited success compared to recent deep generative modeling in visual domain. This paper aims at generating plausible text sentences, whose attributes are controlled by learning disentangled latent representations with designated semantics. We propose a new neural generative model which combines variational auto-encoders (VAEs) and holistic attribute discriminators for effective imposition of semantic structures. The model can alternatively be seen as enhancing VAEs with the wake-sleep algorithm for leveraging fake samples as extra training data. With differentiable approximation to discrete text samples, explicit constraints on independent attribute controls, and efficient collaborative learning of generator and discriminators, our model learns interpretable representations from even only word annotations, and produces short sentences with desired attributes of sentiment and tenses. Quantitative experiments using trained classifiers as evaluators validate the accuracy of sentence and attribute generation.

Author Information

Zhiting Hu (Carnegie Mellon University)
Zichao Yang (Carnegie Mellon University)
Xiaodan Liang (Carnegie Mellon University)
Russ Salakhutdinov (Carnegie Mellen University)
Eric Xing (Carnegie Mellon University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors