Frequency-Guided Expert Modulation for Noisy-Label Facial Expression Recognition
Journal of Imaging · 03.08.2026 · 12 (8) · 350
DOI: 10.3390/jimaging12080350 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Imaging, online erschienen am 03.08.2026. Verfasst von Miaomiao Zhang und 2 weiteren Autor:innen. Beteiligte Einrichtung: Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053, China. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Facial expression recognition in the wild is challenged by both noisy supervision and degraded visual evidence: subtle expression cues must be interpreted under blur, contrast changes, image noise, and annotator disagreement. Existing noisy-label FER methods mainly regulate samples, labels, or attention, while frequency information is rarely used to adapt the semantic representation itself. We propose Frequency-Guided Expert Modulation (FARM-FER), which treats local and global frequency descriptors as a control signal rather than an additional classifier input. A joint Haar-DWT and radial-FFT context guides soft routing among nonlinear experts and channel-wise affine recalibration of the semantic feature, while a learned gate combines the two corrections before a lightweight classifier predicts the expression from the refined representation. Across RAF-DB, FER+, and AffectNet under symmetric label noise, with additional evaluations under class-dependent label noise on RAF-DB and native crowd-label ambiguity on FER+, FARM-FER consistently improves matched baselines. At 30% symmetric noise, FARM-FER reaches 89.18% accuracy on RAF-DB, with a 1.6% performance gain over the matched Swin-Tiny baseline; the gains also hold in a controlled ResNet18 reimplementation and in class-sensitive AffectNet evaluation. Measured cost analyses show only modest parameter and FLOP overhead, supporting a lightweight yet effective design in terms of model size and arithmetic cost for noisy-label FER.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053, China.
- School of Computing and Data Science, The University of Hong Kong, Hong Kong SAR, China.
- Hong Kong Generative AI Research and Development Center, The Hong Kong University of Science and Technology, Hong Kong SAR, China.
- The Hong Kong University of Science and Technology, Hong Kong SAR, China.
ORCID-Profile: Meng Lou, Linwei Chen
Zitieren
APA
Zhang, M., Lou, M., & Chen, L. (2026). Frequency-Guided Expert Modulation for Noisy-Label Facial Expression Recognition. Journal of Imaging, 12(8), 350. https://doi.org/10.3390/jimaging12080350
Vancouver
Zhang M, Lou M, Chen L. Frequency-Guided Expert Modulation for Noisy-Label Facial Expression Recognition. Journal of Imaging. 2026;12(8):350. doi:10.3390/jimaging12080350
BibTeX
@article{Zhang2026,
author = {Zhang, Miaomiao and Lou, Meng and Chen, Linwei},
title = {Frequency-Guided Expert Modulation for Noisy-Label Facial Expression Recognition},
journal = {Journal of Imaging},
year = {2026},
volume = {12},
number = {8},
pages = {350},
doi = {10.3390/jimaging12080350}
}
EUVIMED