Jurassic World Remake: Bringing Ancient Fossils Back to Life via Zero-Shot Long Image-to-Image Translation

TL;DR

This paper introduces a zero-shot image-to-image translation method using text-guided latent diffusion models to translate between skulls and animals, enabling applications in paleontology and other fields.

Contribution

The paper proposes Revive-2I, a novel benchmark model for large domain gap translation using guided diffusion and text prompts, outperforming traditional GANs.

Findings

Guided diffusion models are effective for large domain gap translation.

Prompting with text provides scalable and flexible target domain guidance.

Traditional GANs fail at large domain gap translation.

Abstract

With a strong understanding of the target domain from natural language, we produce promising results in translating across large domain gaps and bringing skeletons back to life. In this work, we use text-guided latent diffusion models for zero-shot image-to-image translation (I2I) across large domain gaps (longI2I), where large amounts of new visual features and new geometry need to be generated to enter the target domain. Being able to perform translations across large domain gaps has a wide variety of real-world applications in criminology, astrology, environmental conservation, and paleontology. In this work, we introduce a new task Skull2Animal for translating between skulls and living animals. On this task, we find that unguided Generative Adversarial Networks (GANs) are not capable of translating across large domain gaps. Instead of these traditional I2I methods, we explore the use of guided diffusion and image editing models and provide a new benchmark model, Revive-2I, capable of performing zero-shot I2I via text-prompting latent diffusion models. We find that guidance is necessary for longI2I because, to bridge the large domain gap, prior knowledge about the target domain is needed. In addition, we find that prompting provides the best and most scalable information about the target domain as classifier-guided diffusion models require retraining for specific use cases and lack stronger constraints on the target domain because of the wide variety of images they are trained on.