English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Scene flow estimation is a long-standing problem in computer vision, wherethe goal is to find the 3D motion of a scene from its consecutive observations.Recently, there have been efforts to compute the scene flow from 3D pointclouds. A common approach is to train a regression model that consumes sourceand target point clouds and outputs the per-point translation vector. Analternative is to learn point matches between the point clouds concurrentlywith regressing a refinement of the initial correspondence flow. In both cases,the learning task is very challenging since the flow regression is done in thefree 3D space, and a typical solution is to resort to a large annotatedsynthetic dataset. We introduce SCOOP, a new method for scene flow estimationthat can be learned on a small amount of data without employing ground-truthflow supervision. In contrast to previous work, we train a pure correspondencemodel focused on learning point feature representation and initialize the flowas the difference between a source point and its softly corresponding targetpoint. Then, in the run-time phase, we directly optimize a flow refinementcomponent with a self-supervised objective, which leads to a coherent andaccurate flow field between the point clouds. Experiments on widespreaddatasets demonstrate the performance gains achieved by our method compared toexisting leading techniques while using a fraction of the training data. Ourcode is publicly available at https://github.com/itailang/SCOOP.

SCOOP: Self-Supervised Correspondence and Optimization-Based Scene Flow