Be-Your-Outpainter: Mastering Video Outpainting through Input-Specific Adaptation

TL;DR

MOTIA is introduced, a diffusion-based pipeline that leverages both the intrinsic data-specific patterns of the source video and the image/video generative prior for effective outpainting, outperforming existing state-of-the-art methods in widely recognized benchmarks.

Abstract

Video outpainting is a challenging task, aiming at generating video content outside the viewport of the input video while maintaining inter-frame and intra-frame consistency. Existing methods fall short in either generation quality or flexibility. We introduce MOTIA Mastering Video Outpainting Through Input-Specific Adaptation, a diffusion-based pipeline that leverages both the intrinsic data-specific patterns of the source video and the image/video generative prior for effective outpainting. MOTIA comprises two main phases: input-specific adaptation and pattern-aware outpainting. The input-specific adaptation phase involves conducting efficient and effective pseudo outpainting learning on the single-shot source video. This process encourages the model to identify and learn patterns within the source video, as well as bridging the gap between standard generative processes and outpainting. The subsequent phase, pattern-aware outpainting, is dedicated to the generalization of these learned patterns to generate outpainting outcomes. Additional strategies including spatial-aware insertion and noise travel are proposed to better leverage the diffusion model's generative prior and the acquired video patterns from source videos. Extensive evaluations underscore MOTIA's superiority, outperforming existing state-of-the-art methods in widely recognized benchmarks. Notably, these advancements are achieved without necessitating extensive, task-specific tuning.

Figures (14)

• Figure 1: MOTIA is a high-quality flexible video outpainting pipeline, leveraging the intrinsic data-specific patterns of source videos and image/video generative prior for state-of-the-art performance. Quantitative metric improvement of MOTIA is significant (Table \ref{['tab:quant']}).
• Figure 2: Failure example of previous methods. Many previous methods including the intensively trained models on video outpainting still might suffer from generation failure, that the model simply generates blurred corners. MOTIA never encounters this failure.
• Figure 3: Workflow of MOTIA. Blue lines represent the workflow of input-specific adaptation, and green lines represent the workflow of pattern-aware outpainting.
• Figure 4: Sample results of quantitative experiments. All videos are outpainted with a horizontal mask ratio of 0.66. Contents outside the yellow lines are outpainted by MOTIA.
• Figure 5: Spatial-aware insertion scales the insertion weights of adapters for better leveraging of learned patterns and generative prior.