Imagine This! Scripts to Compositions to Videos

TL;DR

This paper introduces CRAFT, a model that learns to generate realistic scene videos from natural language captions by predicting layouts, retrieving relevant segments, and fusing them, advancing video synthesis from descriptive text.

Contribution

The paper presents CRAFT, a novel model that explicitly predicts scene layouts, retrieves spatio-temporal segments, and fuses them to generate videos from captions, with sequential training and compositional learning.

Findings

CRAFT outperforms pixel generation methods in semantic fidelity and visual quality.

It generalizes well to unseen captions and video databases.

Demonstrated on the new FLINTSTONES dataset with over 25,000 videos.

Abstract

Imagining a scene described in natural language with realistic layout and appearance of entities is the ultimate test of spatial, visual, and semantic world knowledge. Towards this goal, we present the Composition, Retrieval, and Fusion Network (CRAFT), a model capable of learning this knowledge from video-caption data and applying it while generating videos from novel captions. CRAFT explicitly predicts a temporal-layout of mentioned entities (characters and objects), retrieves spatio-temporal entity segments from a video database and fuses them to generate scene videos. Our contributions include sequential training of components of CRAFT while jointly modeling layout and appearances, and losses that encourage learning compositional representations for retrieval. We evaluate CRAFT on semantic fidelity to caption, composition consistency, and visual quality. CRAFT outperforms direct pixel generation approaches and generalizes well to unseen captions and to unseen video databases with no text annotations. We demonstrate CRAFT on FLINTSTONES, a new richly annotated video-caption dataset with over 25000 videos. For a glimpse of videos generated by CRAFT, see https://youtu.be/688Vv86n0z8.