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

Estimating the 3D motion of points in a scene, known as scene flow, is a coreproblem in computer vision. Traditional learning-based methods designed tolearn end-to-end 3D flow often suffer from poor generalization. Here we presenta recurrent architecture that learns a single step of an unrolled iterativealignment procedure for refining scene flow predictions. Inspired by classicalalgorithms, we demonstrate iterative convergence toward the solution usingstrong regularization. The proposed method can handle sizeable temporaldeformations and suggests a slimmer architecture than competitive all-to-allcorrelation approaches. Trained on FlyingThings3D synthetic data only, ournetwork successfully generalizes to real scans, outperforming all existingmethods by a large margin on the KITTI self-supervised benchmark.

FlowStep3D: Model Unrolling for Self-Supervised Scene Flow Estimation