EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

FreqMambaGAN: A Frequency-Decoupled Mamba-Enhanced CycleGAN for Underwater Image Enhancement

Baojiang Ye, Haifeng Wang, Wenbin Wang, Tianyi Wang

Journal of Marine Science and Engineering · 03.06.2026 · 14 (11) · 1050

DOI: 10.3390/jmse14111050 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Marine Science and Engineering, online erschienen am 03.06.2026. Verfasst von Baojiang Ye und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Department of Science and Technology of Hainan Province, Sanya Science and Technology Bureau, Department of Education of Hainan Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Underwater images often suffer from color cast, low contrast, scattering-induced haze, and texture degradation, which limit the performance of underwater visual perception systems. To address these problems, this study proposes FreqMambaGAN, a frequency-decoupled selective state-space cycle-adversarial network for underwater image enhancement. The proposed method is built upon a CycleGAN-style bidirectional translation framework and introduces a frequency-decoupled Mamba generator to separately model low-frequency color and illumination information and high-frequency texture and edge details. In addition, Efficient Mamba Blocks are embedded into the generator and discriminator to enhance long-range dependency modeling with linear computational complexity. Skip-attention connections are further adopted to preserve shallow spatial details during reconstruction. To improve training stability and imaging plausibility, a multi-stage training strategy is designed by combining supervised warm-up, unpaired cycle-adversarial learning, perceptual regularization, total variation smoothing, and a lightweight physics-inspired consistency constraint based on dark-channel and underwater image-formation priors. Experiments on public underwater image enhancement datasets demonstrate that FreqMambaGAN achieves competitive quantitative performance and visually improved enhancement results in terms of color correction, contrast restoration, haze suppression, and structural preservation. These results indicate that integrating frequency-domain decomposition with selective state-space modeling is effective for underwater image enhancement.

Förderung

  • Department of Science and Technology of Hainan Province
  • Sanya Science and Technology Bureau
  • Department of Education of Hainan Province
  • Hainan Tropical Ocean University
  • Yazhou Bay Innovation Institute, Hainan Tropical Ocean University

Zitieren

APA

Ye, B., Wang, H., Wang, W., & Wang, T. (2026). FreqMambaGAN: A Frequency-Decoupled Mamba-Enhanced CycleGAN for Underwater Image Enhancement. Journal of Marine Science and Engineering, 14(11), 1050. https://doi.org/10.3390/jmse14111050

Vancouver

Ye B, Wang H, Wang W, Wang T. FreqMambaGAN: A Frequency-Decoupled Mamba-Enhanced CycleGAN for Underwater Image Enhancement. Journal of Marine Science and Engineering. 2026;14(11):1050. doi:10.3390/jmse14111050

BibTeX
@article{Ye2026,
  author = {Ye, Baojiang and Wang, Haifeng and Wang, Wenbin and Wang, Tianyi},
  title = {FreqMambaGAN: A Frequency-Decoupled Mamba-Enhanced CycleGAN for Underwater Image Enhancement},
  journal = {Journal of Marine Science and Engineering},
  year = {2026},
  volume = {14},
  number = {11},
  pages = {1050},
  doi = {10.3390/jmse14111050}
}
Kostenlos starten →