Stochastic Dynamics for Video Infilling

TL;DR

This paper presents SDVI, a stochastic dynamics framework for infilling long video intervals with coherent and realistic frame sequences, addressing a different challenge than traditional video interpolation.

Contribution

Introduces a novel stochastic dynamics approach for long-interval video infilling, combining constraint propagation and stochastic sampling for coherent frame generation.

Findings

Generates clear, content-varied frame sequences.

Produces realistic, smoothly transitioning motions.

Outperforms existing methods in coherence and realism.

Abstract

In this paper, we introduce a stochastic dynamics video infilling (SDVI) framework to generate frames between long intervals in a video. Our task differs from video interpolation which aims to produce transitional frames for a short interval between every two frames and increase the temporal resolution. Our task, namely video infilling, however, aims to infill long intervals with plausible frame sequences. Our framework models the infilling as a constrained stochastic generation process and sequentially samples dynamics from the inferred distribution. SDVI consists of two parts: (1) a bi-directional constraint propagation module to guarantee the spatial-temporal coherence among frames, (2) a stochastic sampling process to generate dynamics from the inferred distributions. Experimental results show that SDVI can generate clear frame sequences with varying contents. Moreover, motions in the generated sequence are realistic and able to transfer smoothly from the given start frame to the terminal frame. Our project site is https://xharlie.github.io/projects/project_sites/SDVI/video_results.html