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

Recent advances of monocular 3D object detection facilitate the 3Dmulti-object tracking task based on low-cost camera sensors. In this paper, wefind that the motion cue of objects along different time frames is critical in3D multi-object tracking, which is less explored in existing monocular-basedapproaches. In this paper, we propose a motion-aware framework for monocular 3DMOT. To this end, we propose MoMA-M3T, a framework that mainly consists ofthree motion-aware components. First, we represent the possible movement of anobject related to all object tracklets in the feature space as its motionfeatures. Then, we further model the historical object tracklet along the timeframe in a spatial-temporal perspective via a motion transformer. Finally, wepropose a motion-aware matching module to associate historical object trackletsand current observations as final tracking results. We conduct extensiveexperiments on the nuScenes and KITTI datasets to demonstrate that our MoMA-M3Tachieves competitive performance against state-of-the-art methods. Moreover,the proposed tracker is flexible and can be easily plugged into existingimage-based 3D object detectors without re-training. Code and models areavailable at https://github.com/kuanchihhuang/MoMA-M3T.

Delving into Motion-Aware Matching for Monocular 3D Object Tracking