StoryGAN: A Sequential Conditional GAN for Story Visualization

TL;DR

StoryGAN introduces a novel sequential conditional GAN model for story visualization, generating consistent image sequences from multi-sentence paragraphs, emphasizing global scene and character coherence over frame continuity.

Contribution

The paper presents a new model, StoryGAN, with a deep Context Encoder and dual discriminators, specifically designed for global consistency in story visualization tasks.

Findings

StoryGAN outperforms existing models in image quality.

It achieves higher contextual consistency in generated sequences.

Human evaluations favor StoryGAN's outputs.

Abstract

We propose a new task, called Story Visualization. Given a multi-sentence paragraph, the story is visualized by generating a sequence of images, one for each sentence. In contrast to video generation, story visualization focuses less on the continuity in generated images (frames), but more on the global consistency across dynamic scenes and characters -- a challenge that has not been addressed by any single-image or video generation methods. We therefore propose a new story-to-image-sequence generation model, StoryGAN, based on the sequential conditional GAN framework. Our model is unique in that it consists of a deep Context Encoder that dynamically tracks the story flow, and two discriminators at the story and image levels, to enhance the image quality and the consistency of the generated sequences. To evaluate the model, we modified existing datasets to create the CLEVR-SV and Pororo-SV datasets. Empirically, StoryGAN outperforms state-of-the-art models in image quality, contextual consistency metrics, and human evaluation.