EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Lightweight Architecture Optimization for Car Logo Detection Based on Visual State Space Models

Jingwen Gong, Junxing Zhang iD, Chunjuan Bo

Sensors · 04.11.2026 · 26 (21) · 6973

DOI: 10.3390/s26216973 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 04.11.2026. Verfasst von Jingwen Gong und 2 weiteren Autor:innen. Beteiligte Einrichtung: College of Information and Communication Engineering, Dalian Minzu University, Dalian 116600, China. Die Arbeit nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Liaoning Provincial Science and Technology Plan Joint Program, Education Reform Research Project of the State Ethnic Affairs Commission of China, Joint Program of Liaoning Provincial Science and Technology Plan.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Car logo detection is a challenging fine-grained perception task characterized by small object scales and high inter-class similarity. In such scenarios, discriminative cues mainly reside in subtle local textures rather than global structures. Although State Space Model (SSM)-based detectors, such as Mamba-YOLO with its ODMamba backbone, provide efficient long-range dependency modeling, they often struggle to maintain spatial alignment and semantic consistency during multi-scale feature fusion. This limitation reduces their effectiveness in fine-grained discrimination. To address these issues, this paper proposes LVSS-Net (Lightweight Visual State Space Network), a unified framework with three key components. First, Multi-Scale Semantic Alignment (MSAA) preserves spatial resolution while enhancing semantic consistency across feature pyramid levels through adaptive normalization and attention-based fusion. Second, Dynamic Content-Aware Reconstruction (DCR) introduces learnable content-adaptive offsets to model variable sampling positions during feature upsampling, mitigating scale-induced information loss and boundary blurring. Third, Global–Local Saliency Modulation (GLSM) combines parameter-free global saliency modeling with local texture enhancement via a lightweight LTE module to capture fine-grained discriminative patterns. Extensive experiments on the VLD-11 fine-grained car logo dataset and the TT100K traffic sign detection dataset demonstrate that LVSS-Net achieves 98.3% mAP@0.5 with only 1.5 M parameters. The proposed method reduces model complexity by 73.5% compared to the Mamba-YOLO baseline while maintaining competitive detection performance. These results demonstrate its effectiveness in resource-constrained fine-grained detection scenarios.

Schlagwörter

Stichwörter der Autor:innen

Mamba-YOLOcar logo detectiondynamic samplingfeature enhancementlightweight object detectionstate space model (SSM)

Autor:innen und Institutionen

  • College of Information and Communication Engineering, Dalian Minzu University, Dalian 116600, China.
  • College of Computer Science and Engineering, Dalian Minzu University, Dalian 116600, China.

ORCID-Profile: Junxing Zhang

Förderung

  • Liaoning Provincial Science and Technology Plan Joint Program
  • Education Reform Research Project of the State Ethnic Affairs Commission of China
  • Joint Program of Liaoning Provincial Science and Technology Plan

Zitieren

APA

Gong, J., Zhang, J., & Bo, C. (2026). Lightweight Architecture Optimization for Car Logo Detection Based on Visual State Space Models. Sensors, 26(21), 6973. https://doi.org/10.3390/s26216973

Vancouver

Gong J, Zhang J, Bo C. Lightweight Architecture Optimization for Car Logo Detection Based on Visual State Space Models. Sensors. 2026;26(21):6973. doi:10.3390/s26216973

BibTeX
@article{Gong2026,
  author = {Gong, Jingwen and Zhang, Junxing and Bo, Chunjuan},
  title = {Lightweight Architecture Optimization for Car Logo Detection Based on Visual State Space Models},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {21},
  pages = {6973},
  doi = {10.3390/s26216973}
}
Kostenlos starten →