arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.11280cs.CV
收藏

关节动画的运动表示

Motion Representations for Articulated Animation

University of Trento(特伦托大学);Snap Inc(Snap公司)

Aliaksandr Siarohin, Oliver J. Woodford, Jian Ren, Menglei Chai, Sergey Tulyakov

摘要:

我们提出了新的运动表示动画连接的对象组成的不同部分。在一个完全无监督的方式,我们的方法识别对象的部分,跟踪他们在驾驶视频,并推断他们的运动,通过考虑他们的主轴。与以前的基于关键点的工作相比,我们的方法提取有意义的和一致的区域,描述位置,形状和姿势。这些区域对应于语义上相关且不同的对象部分,其在驾驶视频的帧中更容易被检测到。为了强制前景与背景的解耦,我们用额外的仿射变换来建模非对象相关的全局运动。为了方便动画和防止泄漏的形状的驱动对象,我们解开形状和姿态的区域空间中的对象。我们的模型可以制作各种对象的动画,在现有的基准测试中大大超过了以前的方法。我们提出了一个具有挑战性的高分辨率视频新基准,并表明当考虑铰接对象时,这种改进特别明显,达到96.6%的用户偏好与最先进的水平。

英文摘要:

We propose novel motion representations for animating articulated objects consisting of distinct parts. In a completely unsupervised manner, our method identifies object parts, tracks them in a driving video, and infers their motions by considering their principal axes. In contrast to the previous keypoint-based works, our method extracts meaningful and consistent regions, describing locations, shape, and pose. The regions correspond to semantically relevant and distinct object parts, that are more easily detected in frames of the driving video. To force decoupling of foreground from background, we model non-object related global motion with an additional affine transformation. To facilitate animation and prevent the leakage of the shape of the driving object, we disentangle shape and pose of objects in the region space. Our model can animate a variety of objects, surpassing previous methods by a large margin on existing benchmarks. We present a challenging new benchmark with high-resolution videos and show that the improvement is particularly pronounced when articulated objects are considered, reaching 96.6% user preference vs. the state of the art.

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。

↑