DynVFX: Augmenting Real Videos with Dynamic Content

TL;DR

DynVFX is a zero-shot framework that seamlessly augments real videos with dynamic content based on simple text instructions, integrating new objects or effects naturally into existing scenes.

Contribution

It introduces a novel inference-based method that manipulates attention features for realistic, automated video augmentation without training, leveraging pre-trained text-to-video and vision-language models.

Findings

Effective augmentation of real videos with diverse dynamic content

Seamless integration accounting for camera motion and occlusions

Automated process requiring only simple user instructions

Abstract

We present a method for augmenting real-world videos with newly generated dynamic content. Given an input video and a simple user-provided text instruction describing the desired content, our method synthesizes dynamic objects or complex scene effects that naturally interact with the existing scene over time. The position, appearance, and motion of the new content are seamlessly integrated into the original footage while accounting for camera motion, occlusions, and interactions with other dynamic objects in the scene, resulting in a cohesive and realistic output video. We achieve this via a zero-shot, training-free framework that harnesses a pre-trained text-to-video diffusion transformer to synthesize the new content and a pre-trained vision-language model to envision the augmented scene in detail. Specifically, we introduce a novel inference-based method that manipulates features within the attention mechanism, enabling accurate localization and seamless integration of the new content while preserving the integrity of the original scene. Our method is fully automated, requiring only a simple user instruction. We demonstrate its effectiveness on a wide range of edits applied to real-world videos, encompassing diverse objects and scenarios involving both camera and object motion.