Debiased Multiplex Tokenizer for Efficient Map-Free Visual Relocalization
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (12) · 10145-10153
DOI: 10.1609/aaai.v40i12.37982 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Wenshuai Wang und 4 weiteren Autor:innen.
Abstract (Originalsprache)
Image-based feature representation plays a critical role in visual localization, enabling robots to estimate their position and orientation in GPS-denied environments. However, this task is often undermined by significant variations in camera viewpoints and scene appearances. Recently, map-free visual relocalization (MFVR) has emerged as a promising paradigm due to its compatibility with lightweight deployment and privacy isolation on mobile devices. In this paper, we propose the Debiased Multiplex Tokenizer (DeMT) as a novel method for versatile and efficient MFVR. Specifically, DeMT performs relative pose regression through an integrated framework built upon a pretrained vision Mamba encoder, comprising three key modules: First, Multiplex Interactive Tokenization yields robust image tokens with non-local affinities and cross-domain descriptions; Second, Debiased Anchor Registration facilitates anchor token matching through proximity graph retrieval and causal pointer attribution; Third, Geometry-Informed Pose Regression empowers multi-layer perceptrons with a gating mechanism and spectral normalization to support both pair-wise and multi-view modes. Extensive evaluations across nine public datasets demonstrate that DeMT substantially outperforms existing baselines and ablation variants in diverse indoor and outdoor environments.
Zitieren
APA
Wang, W., Liu, H., Li, S., Jiang, P., & Ding, R. (2026). Debiased Multiplex Tokenizer for Efficient Map-Free Visual Relocalization. Proceedings of the AAAI Conference on Artificial Intelligence, 40(12), 10145-10153. https://doi.org/10.1609/aaai.v40i12.37982
Vancouver
Wang W, Liu H, Li S, Jiang P, Ding R. Debiased Multiplex Tokenizer for Efficient Map-Free Visual Relocalization. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(12):10145-10153. doi:10.1609/aaai.v40i12.37982
BibTeX
@article{Wang2026,
author = {Wang, Wenshuai and Liu, Hong and Li, Shengquan and Jiang, Peifeng and Ding, Runwei},
title = {Debiased Multiplex Tokenizer for Efficient Map-Free Visual Relocalization},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {12},
pages = {10145--10153},
doi = {10.1609/aaai.v40i12.37982}
}
EUVIMED