EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Integrating Hybrid Attention Mechanisms into CNN-Based Architectures to Enhance Image Classification and Interpretability

Alidor M. Mbayandjambe, Selain K. Kasereka, Darren Kevin T. Nguemdjom, Petro M. Tshakwanda, Milena Savova-Mratsenkova, Tasho Tashev

Machine Learning and Knowledge Extraction · 25.05.2026 · 8 (6) · 143

DOI: 10.3390/make8060143 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Machine Learning and Knowledge Extraction, online erschienen am 25.05.2026. Verfasst von Alidor M. Mbayandjambe und 5 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Regional Development Fund.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Integrating complementary attention mechanisms into standard Convolutional Neural Networks (CNNs) is a promising strategy for improving feature discrimination without substantial computational overhead. This paper presents a controlled empirical study of a hybrid attention module that combines Squeeze-and-Excitation Networks (SENet) and the Convolutional Block Attention Module (CBAM) through an adaptive element-wise summation with a learnable weighting parameter α and a residual connection. This work contributes a systematic and statistically rigorous evaluation of attention fusion across four CNN backbones (ResNet18, VGG16, AlexNet, and SqueezeNet) on the CIFAR-10 benchmark at 32×32 resolution. All models were trained from scratch under a deliberately conservative protocol (50 epochs, no pretrained weights, standard augmentation) to isolate the incremental effect of attention mechanisms under controlled conditions. Under this protocol, the hybrid SENet+CBAM configuration achieves statistically significant accuracy improvements over the corresponding baselines (p

Förderung

  • European Regional Development Fund

Zitieren

APA

Mbayandjambe, A. M., Kasereka, S. K., Nguemdjom, D. K. T., Tshakwanda, P. M., Savova-Mratsenkova, M., & Tashev, T. (2026). Integrating Hybrid Attention Mechanisms into CNN-Based Architectures to Enhance Image Classification and Interpretability. Machine Learning and Knowledge Extraction, 8(6), 143. https://doi.org/10.3390/make8060143

Vancouver

Mbayandjambe AM, Kasereka SK, Nguemdjom DKT, Tshakwanda PM, Savova-Mratsenkova M, Tashev T. Integrating Hybrid Attention Mechanisms into CNN-Based Architectures to Enhance Image Classification and Interpretability. Machine Learning and Knowledge Extraction. 2026;8(6):143. doi:10.3390/make8060143

BibTeX
@article{Mbayandjambe2026,
  author = {Mbayandjambe, Alidor M. and Kasereka, Selain K. and Nguemdjom, Darren Kevin T. and Tshakwanda, Petro M. and Savova-Mratsenkova, Milena and Tashev, Tasho},
  title = {Integrating Hybrid Attention Mechanisms into CNN-Based Architectures to Enhance Image Classification and Interpretability},
  journal = {Machine Learning and Knowledge Extraction},
  year = {2026},
  volume = {8},
  number = {6},
  pages = {143},
  doi = {10.3390/make8060143}
}
Kostenlos starten →