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

A key challenge for autonomous vehicles is to navigate in unseen dynamicenvironments. Separating moving objects from static ones is essential fornavigation, pose estimation, and understanding how other traffic participantsare likely to move in the near future. In this work, we tackle the problem ofdistinguishing 3D LiDAR points that belong to currently moving objects, likewalking pedestrians or driving cars, from points that are obtained fromnon-moving objects, like walls but also parked cars. Our approach takes asequence of observed LiDAR scans and turns them into a voxelized sparse 4Dpoint cloud. We apply computationally efficient sparse 4D convolutions tojointly extract spatial and temporal features and predict moving objectconfidence scores for all points in the sequence. We develop a receding horizonstrategy that allows us to predict moving objects online and to refinepredictions on the go based on new observations. We use a binary Bayes filterto recursively integrate new predictions of a scan resulting in more robustestimation. We evaluate our approach on the SemanticKITTI moving objectsegmentation challenge and show more accurate predictions than existingmethods. Since our approach only operates on the geometric information of pointclouds over time, it generalizes well to new, unseen environments, which weevaluate on the Apollo dataset.

Receding Moving Object Segmentation in 3D LiDAR Data Using Sparse 4D Convolutions