EVCC: Enhanced Vision Transformer-ConvNeXt-CoAtNet Fusion for Classification
By: Kazi Reyazul Hasan , Md Nafiu Rahman , Wasif Jalal and more
Potential Business Impact:
Makes AI see better, using less computer power.
Hybrid vision architectures combining Transformers and CNNs have significantly advanced image classification, but they usually do so at significant computational cost. We introduce EVCC (Enhanced Vision Transformer-ConvNeXt-CoAtNet), a novel multi-branch architecture integrating the Vision Transformer, lightweight ConvNeXt, and CoAtNet through key innovations: (1) adaptive token pruning with information preservation, (2) gated bidirectional cross-attention for enhanced feature refinement, (3) auxiliary classification heads for multi-task learning, and (4) a dynamic router gate employing context-aware confidence-driven weighting. Experiments across the CIFAR-100, Tobacco3482, CelebA, and Brain Cancer datasets demonstrate EVCC's superiority over powerful models like DeiT-Base, MaxViT-Base, and CrossViT-Base by consistently achieving state-of-the-art accuracy with improvements of up to 2 percentage points, while reducing FLOPs by 25 to 35%. Our adaptive architecture adjusts computational demands to deployment needs by dynamically reducing token count, efficiently balancing the accuracy-efficiency trade-off while combining global context, local details, and hierarchical features for real-world applications. The source code of our implementation is available at https://anonymous.4open.science/r/EVCC.
Similar Papers
ECViT: Efficient Convolutional Vision Transformer with Local-Attention and Multi-scale Stages
CV and Pattern Recognition
Makes AI see pictures faster and better.
CoCAViT: Compact Vision Transformer with Robust Global Coordination
CV and Pattern Recognition
Makes small computer vision models work better everywhere.
CoAtNeXt:An Attention-Enhanced ConvNeXtV2-Transformer Hybrid Model for Gastric Tissue Classification
CV and Pattern Recognition
Helps doctors find stomach sickness faster.