A pinboard by Ofir Press

Finishing M.Sc. and starting Ph.D., Tel-Aviv University

Pinboard Summary

Generative Adversarial Networks (GANs) have shown great promise recently in image generation. Training GANs for text generation has proven to be more difficult, because of the non-differentiable nature of generating text with recurrent neural networks. Consequently, past work has either resorted to pre-training with maximum-likelihood or used convolutional networks for generation. In this work, we show that recurrent neural networks can be trained to generate text with GANs from scratch by employing curriculum learning, slowly increasing the length of the generated text, and by training the RNN simultaneously to generate sequences of different lengths. We show that this approach vastly improves the quality of generated sequences compared to the convolutional baseline.

11 items pinned

Recurrent Topic-Transition GAN for Visual Paragraph Generation

Abstract: A natural image usually conveys rich semantic content and can be viewed from different angles. Existing image description methods are largely restricted by small sets of biased visual paragraph annotations, and fail to cover rich underlying semantics. In this paper, we investigate a semi-supervised paragraph generative framework that is able to synthesize diverse and semantically coherent paragraph descriptions by reasoning over local semantic regions and exploiting linguistic knowledge. The proposed Recurrent Topic-Transition Generative Adversarial Network (RTT-GAN) builds an adversarial framework between a structured paragraph generator and multi-level paragraph discriminators. The paragraph generator generates sentences recurrently by incorporating region-based visual and language attention mechanisms at each step. The quality of generated paragraph sentences is assessed by multi-level adversarial discriminators from two aspects, namely, plausibility at sentence level and topic-transition coherence at paragraph level. The joint adversarial training of RTT-GAN drives the model to generate realistic paragraphs with smooth logical transition between sentence topics. Extensive quantitative experiments on image and video paragraph datasets demonstrate the effectiveness of our RTT-GAN in both supervised and semi-supervised settings. Qualitative results on telling diverse stories for an image also verify the interpretability of RTT-GAN.

Pub.: 20 Mar '17, Pinned: 15 Jun '17