arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.01700cs.CVcs.MM
收藏

视频扩散模型能否预测过去帧?双向循环一致性用于可逆插值

Can Video Diffusion Models Predict Past Frames? Bidirectional Cycle Consistency for Reversible Interpolation

School of Software Engineering, Xi'an Jiaotong University(西安交通大学软件工程学院);School of Computer and Information Science, Hefei University of Technology(合肥工业大学计算机与信息科学学院);Faculty of Science and Technology, University of Macau(澳门大学科技学院)

Lingyu Liu, Yaxiong Wang, Li Zhu, Zhedong Zheng

AI总结:

本文提出双向循环一致性框架,通过双向生成轨迹对称性提升视频插值的时序一致性,实验表明在37帧和73帧任务中取得最佳性能。

摘要:

视频帧插值旨在在给定端点之间合成逼真中间帧,同时遵循特定运动语义。尽管生成模型在视觉保真度上有所改进,但其主要以单向方式工作,缺乏自我验证时间一致性机制。受自监督学习中时间循环一致性的启发,我们提出了一种新的双向框架,强制正向和反向生成轨迹的对称性。我们的方法引入可学习的方向标记,以显式地将共享主干网络条件化为时间方向,使模型能够在单一统一架构中联合优化正向合成和反向重建。这种循环一致的监督作用作为一种强大的正则化器,确保生成的运动路径在逻辑上是可逆的。此外,我们采用课程学习策略,从短序列逐步训练模型,稳定不同持续时间的动力学。关键的是,我们的循环约束仅在训练期间应用;推理需要单次前向传递,保持基础模型的高效率。广泛的实验表明,我们的方法在37帧和73帧任务中在成像质量、运动平滑度和动态控制方面均达到最佳性能,优于强基线,且无额外计算开销。

英文摘要:

Video frame interpolation aims to synthesize realistic intermediate frames between given endpoints while adhering to specific motion semantics. While recent generative models have improved visual fidelity, they predominantly operate in a unidirectional manner, lacking mechanisms to self-verify temporal consistency. This often leads to motion drift, directional ambiguity, and boundary misalignment, especially in long-range sequences. Inspired by the principle of temporal cycle-consistency in self-supervised learning, we propose a novel bidirectional framework that enforces symmetry between forward and backward generation trajectories. Our approach introduces learnable directional tokens to explicitly condition a shared backbone on temporal orientation, enabling the model to jointly optimize forward synthesis and backward reconstruction within a single unified architecture. This cycle-consistent supervision acts as a powerful regularizer, ensuring that generated motion paths are logically reversible. Furthermore, we employ a curriculum learning strategy that progressively trains the model from short to long sequences, stabilizing dynamics across varying durations. Crucially, our cyclic constraints are applied only during training; inference requires a single forward pass, maintaining the high efficiency of the base model. Extensive experiments show that our method achieves state-of-the-art performance in imaging quality, motion smoothness, and dynamic control on both 37-frame and 73-frame tasks, outperforming strong baselines while incurring no additional computational overhead.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑