RyNet: Multi-Level Attention for Rich and Complementary Feature Representation in CNNs
Open Access
- 24 December 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 5358-5373
Abstract
Convolutional neural networks (CNNs) excel at extracting spatial and channel-wise features, but reliance on final-layer activations often underutilizes intermediate representations and weakens hierarchical cues. We propose RyNet, a theoretically grounded, modular, and CNN-agnostic dual-attention framework that integrates dynamic channel refinement with multi-level feature fusion in a single backbone. RyNet aggregates low-, mid-, and high-level feature maps, applies channel squeezing with learnable channel-wise attention scores to promote selective activation, and fuses recalibrated features via a lightweight summation strategy. This yields stage-to-stage coupling with implicit spatial emphasis at negligible computational cost. RyNet achieves hierarchical complementarity without branch duplication or concatenation–projection stacks. Integrated into VGG16 and ResNet18, RyNet improves CIFAR-10 test accuracy from 89.90% to 90.84% (a relative improvement of 1.04%) and from 90.42% to 91.74% (a relative improvement of 1.50%), respectively, On the PlantVillage (tomato) dataset, a custom CNN-based model integrated with RyNet achieves 99.61% accuracy, representing a relative improvement of 1.54%. Under identical datasets and training conditions, RyNet consistently and statistically outperforms SE and CBAM, and beyond this, demonstrates superiority over recent CNN–Transformer hybrids and other state-of-the-art methods. Feature-space analysis reveals a 121.9% increase in inter-class separability, while visualizations and interpretability evaluations confirm the effectiveness of dynamic channel selection under multi-level fusion. These results provide strong theoretical, empirical, and statistical validation for the proposed design. Overall, RyNet emerges as a principled, lightweight, and easily integrable module for enhancing CNN representations in resource-constrained environments, with significant performance gains and improved model explainability.Keywords
This publication has 29 references indexed in Scilit:
- Dynamic attention-based transfer learning strategies for efficient and scalable visual recognition in resource-constrained environmentsApplied Soft Computing, 2025
- End-to-end food condition system via wavelet attention enhanced networkNeurocomputing, 2025
- Binary pooling: A novel approach for local feature extraction in convolutional neural networksNeurocomputing, 2025
- Performance evaluation and optimization of convolutional neural network architectures for Tomato plant disease eleven classes based on augmented leaf images datasetNeural Computing and Applications, 2024
- SparseSwin: Swin transformer with sparse transformer blockNeurocomputing, 2024
- A data efficient transformer based on Swin TransformerThe Visual Computer, 2023
- Res‐SE‐Net: Boosting Performance of ResNets by Enhancing Bridge Connections2021
- Tomato plant disease detection using transfer learning with C-GAN synthetic imagesComputers and Electronics in Agriculture, 2021
- Analysis of the Influence of Stylized-CIFAR10 Dataset on ResNet2020
- Visualizing and Understanding Convolutional Networks2014