Training-Free Semantic Video Composition via Pre-trained Diffusion Model

TL;DR

This paper introduces a training-free method for semantic video composition using a pre-trained diffusion model, effectively handling deep semantic disparities and ensuring visual harmony and coherence across frames.

Contribution

It proposes a novel training-free pipeline with Balanced Partial Inversion and Inter-Frame Augmented attention to improve semantic video composition without additional training.

Findings

Successfully manages broader semantic disparities in videos

Ensures inter-frame coherence and visual harmony

Outperforms existing methods in semantic video composition

Abstract

The video composition task aims to integrate specified foregrounds and backgrounds from different videos into a harmonious composite. Current approaches, predominantly trained on videos with adjusted foreground color and lighting, struggle to address deep semantic disparities beyond superficial adjustments, such as domain gaps. Therefore, we propose a training-free pipeline employing a pre-trained diffusion model imbued with semantic prior knowledge, which can process composite videos with broader semantic disparities. Specifically, we process the video frames in a cascading manner and handle each frame in two processes with the diffusion model. In the inversion process, we propose Balanced Partial Inversion to obtain generation initial points that balance reversibility and modifiability. Then, in the generation process, we further propose Inter-Frame Augmented attention to augment foreground continuity across frames. Experimental results reveal that our pipeline successfully ensures the visual harmony and inter-frame coherence of the outputs, demonstrating efficacy in managing broader semantic disparities.