RegTrack: Simplicity Beneath Complexity in Robust Multi-Modal 3D Multi-Object Tracking
IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026-11 · 48 (11) · 14631-14648
DOI: 10.1109/tpami.2026.3719678 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026-11. Verfasst von Lipeng Gu und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 79 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Shenzhen Science and Technology Program, Guangdong Basic and Applied Basic Research Foundation.
Abstract (Originalsprache)
Existing 3D multi-object tracking (MOT) methods often trade efficiency and generalizability for robustness, as they typically rely on complex association metrics derived from multi-modal architectures or class-specific motion priors. Challenging the common belief that greater complexity necessarily leads to stronger robustness, we propose a robust, efficient, and generalizable method for multi-modal 3D MOT, dubbed RegTrack. Inspired by Yang-Mills gauge theory, RegTrack formulates multi-modal 3D MOT as motion-compensated representation learning. Under this analogy, point-cloud object representations are viewed as matter fields, while inter-frame object motions are regarded as local variations. Geometric cues are modeled as gauge fields to adaptively compensate for such variations, and a pretrained image representation space serves as a globally invariant physical law to guide the compensation process. In this way, the resulting motion-compensated point-cloud representations, viewed as observables, are encouraged to remain consistent for the same object across frames while preserving discriminability among different objects. Their pairwise similarities thus provide a simple yet robust association metric. Specifically, RegTrack is built upon a unified tri-cue encoder (UTEnc), which consists of a local-global point cloud encoder (LG-PEnc), a mixture-of-experts-based geometry encoder (MoE-GEnc), and a frozen image encoder derived from a pretrained vision-language model. LG-PEnc efficiently encodes the spatial-structural information of object point clouds to generate foundational representations. MoE-GEnc interacts with LG-PEnc to model inter-frame geometric relationships and adaptively compensate for motion-induced representation variations without relying on class-specific priors. The frozen image encoder is used only during training to provide a stable representation space for supervising the compensation process, and is discarded during inference. As a result, RegTrack achieves robust, efficient, and generalizable inference using only point-cloud inputs, with merely 2.67 M parameters. Extensive experiments on KITTI and nuScenes demonstrate that RegTrack outperforms its thirty-six competitors.
Förderung
- National Natural Science Foundation of China
- Shenzhen Science and Technology Program
- Guangdong Basic and Applied Basic Research Foundation
Zitieren
APA
Gu, L., Yan, X., Wang, S., & Wei, M. (2026). RegTrack: Simplicity Beneath Complexity in Robust Multi-Modal 3D Multi-Object Tracking. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(11), 14631-14648. https://doi.org/10.1109/tpami.2026.3719678
Vancouver
Gu L, Yan X, Wang S, Wei M. RegTrack: Simplicity Beneath Complexity in Robust Multi-Modal 3D Multi-Object Tracking. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026;48(11):14631-14648. doi:10.1109/tpami.2026.3719678
BibTeX
@article{Gu2026,
author = {Gu, Lipeng and Yan, Xuefeng and Wang, Song and Wei, Mingqiang},
title = {RegTrack: Simplicity Beneath Complexity in Robust Multi-Modal 3D Multi-Object Tracking},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
year = {2026},
volume = {48},
number = {11},
pages = {14631--14648},
doi = {10.1109/tpami.2026.3719678}
}
EUVIMED