Purpose: Generative Engine is a story-writing machine designed to automatically generate synthetic images as you write new words and sentences, for a story creation experience that combines text and images.
Main capability: It can automatically generate synthetic images based on the input text content and dynamically produce images as the writing process unfolds. This experiment uses the AttnGAN model.
How to use: Visit https://experiments.runwayml.com/generative_engine/ page, write new words and sentences on the page, and the system will automatically generate corresponding synthetic images.
Use cases: Suitable for narrative experiments exploring the combination of text and images, and for experiencing the text-to-image generation effects of AttnGAN.
Limitations: This page is a Runway Experiment, and information regarding fees, pricing, supported languages, image specifications, commercial licensing, etc., is not currently provided. The model belongs to Tao Xu et al. and uses the MIT License, copyright 2018.