Learning Stochastic Bridges for Video Object Removal via Video-to-Video Translation

TL;DR

This paper introduces a novel video object removal method using a stochastic bridge model that directly translates input videos to clean versions, leveraging structural priors for more accurate and consistent removal.

Contribution

The paper proposes a stochastic bridge framework for video object removal that directly connects source and target videos, improving guidance and scene consistency over diffusion-based methods.

Findings

Outperforms existing methods in visual quality

Achieves superior temporal consistency

Effectively handles large object removal

Abstract

Existing video object removal methods predominantly rely on diffusion models following a noise-to-data paradigm, where generation starts from uninformative Gaussian noise. This approach discards the rich structural and contextual priors present in the original input video. Consequently, such methods often lack sufficient guidance, leading to incomplete object erasure or the synthesis of implausible content that conflicts with the scene's physical logic. In this paper, we reformulate video object removal as a video-to-video translation task via a stochastic bridge model. Unlike noise-initialized methods, our framework establishes a direct stochastic path from the source video (with objects) to the target video (objects removed). This bridge formulation effectively leverages the input video as a strong structural prior, guiding the model to perform precise removal while ensuring that the filled regions are logically consistent with the surrounding environment. To address the trade-off where strong bridge priors hinder the removal of large objects, we propose a novel adaptive mask modulation strategy. This mechanism dynamically modulates input embeddings based on mask characteristics, balancing background fidelity with generative flexibility. Extensive experiments demonstrate that our approach significantly outperforms existing methods in both visual quality and temporal consistency. The project page is https://bridgeremoval.github.io/.