Timezone: »

CaSPR: Learning Canonical Spatiotemporal Point Cloud Representations
Davis Rempe · Tolga Birdal · Yongheng Zhao · Zan Gojcic · Srinath Sridhar · Leonidas Guibas

Wed Dec 09 07:30 PM -- 07:40 PM (PST) @ Orals & Spotlights: Vision Applications

We propose CaSPR, a method to learn object-centric Canonical Spatiotemporal Point Cloud Representations of dynamically moving or evolving objects. Our goal is to enable information aggregation over time and the interrogation of object state at any spatiotemporal neighborhood in the past, observed or not. Different from previous work, CaSPR learns representations that support spacetime continuity, are robust to variable and irregularly spacetime-sampled point clouds, and generalize to unseen object instances. Our approach divides the problem into two subtasks. First, we explicitly encode time by mapping an input point cloud sequence to a spatiotemporally-canonicalized object space. We then leverage this canonicalization to learn a spatiotemporal latent representation using neural ordinary differential equations and a generative model of dynamically evolving shapes using continuous normalizing flows. We demonstrate the effectiveness of our method on several applications including shape reconstruction, camera pose estimation, continuous spatiotemporal sequence reconstruction, and correspondence estimation from irregularly or intermittently sampled observations.

Author Information

Davis Rempe (Stanford University)
Tolga Birdal (Stanford University)
Yongheng Zhao (University of Padova)
Zan Gojcic (ETH Zürich)
Srinath Sridhar (Brown University)
Leonidas Guibas (stanford.edu)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors