EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

LDR-Net: Landmark-Guided Diverse Regional Representation Learning for Facial Expression Recognition

Yansha Lu, Faliang Chang iD, Chunsheng Liu iD, Hui Liu iD, Yiming Huang iD

Biomimetics · 17.07.2026 · 11 (7) · 503

DOI: 10.3390/biomimetics11070503 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Biomimetics, online erschienen am 17.07.2026. Verfasst von Yansha Lu und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Control Science and Engineering, Shandong University, Jinan 250061, China. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Shandong Outstanding Youth Funding, Taishan Scholars Project Special Funding.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reliable facial expression recognition (FER) is essential for human–computer interaction across scenarios. However, existing methods often overlook spatial structure when seeking region-specific features due to fixed or limited local regions, limiting representation capability under real-world variations. Inspired by the multi-location attention of the human visual system, we propose a Landmark-guided Diverse Regional Representation Network (LDR-Net), using dynamic landmarks as patch centers to preserve structural details while locating expression-critical areas, facilitating effective regional representations for FER. First, a novel Diverse Regional Feature Extraction (DRFE) module operates via complementary operations: landmark-guided cropping for local details and cross-level integration with feature reorganization for holistic aggregation. Second, a novel Diverse Representation Learning (DRL) module is proposed with a collaborative dual-stream mechanism that captures fine-grained local dependencies via Transformers while reinforcing global features through attention-based enhancement, enabling comprehensive feature learning. Finally, a new Hybrid Feature Fusion (HFF) module is proposed for joint decision optimization via a hierarchical hybrid strategy, which aggregates intra-branch predictions followed by weighted branch-level fusion. Experiments on three FER benchmarks (RAF-DB, AffectNet, SFEW) and five occlusion/pose test sets demonstrate that LDR-Net outperforms state-of-the-art methods, while cross-scene validation on KMU-FED confirms its effectiveness in real-world driving.

Schlagwörter

Stichwörter der Autor:innen

diverse regional representationfacial expression recognitionhuman visual systemlandmark-guided representationocclusion-robust and pose-invariant

Autor:innen und Institutionen

  • School of Control Science and Engineering, Shandong University, Jinan 250061, China.

ORCID-Profile: Faliang Chang, Chunsheng Liu, Hui Liu, Yiming Huang

Förderung

  • National Natural Science Foundation of China
  • Shandong Outstanding Youth Funding
  • Taishan Scholars Project Special Funding

Zitieren

APA

Lu, Y., Chang, F., Liu, C., Liu, H., & Huang, Y. (2026). LDR-Net: Landmark-Guided Diverse Regional Representation Learning for Facial Expression Recognition. Biomimetics, 11(7), 503. https://doi.org/10.3390/biomimetics11070503

Vancouver

Lu Y, Chang F, Liu C, Liu H, Huang Y. LDR-Net: Landmark-Guided Diverse Regional Representation Learning for Facial Expression Recognition. Biomimetics. 2026;11(7):503. doi:10.3390/biomimetics11070503

BibTeX
@article{Lu2026,
  author = {Lu, Yansha and Chang, Faliang and Liu, Chunsheng and Liu, Hui and Huang, Yiming},
  title = {LDR-Net: Landmark-Guided Diverse Regional Representation Learning for Facial Expression Recognition},
  journal = {Biomimetics},
  year = {2026},
  volume = {11},
  number = {7},
  pages = {503},
  doi = {10.3390/biomimetics11070503}
}
Kostenlos starten →