2025/05/28 by Xuyang Zhang, Zhang, Xuyang, Xi Zhang +7
Computer Science · #Advanced Neural Network Applications #Adversarial Robustness in Machine Learning #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Stochastic Gradient Optimization Techniques
paper · pdf · doi:10.48550/arxiv.2505.22226
openalex publication_date 2025/05/28 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Recent theoretical advances reveal that the Hadamard product induces nonlinear representations and implicit high-dimensional mappings for the field of deep learning, yet their practical deployment in resource-constrained vision models remains largely unexplored. To address this gap, we introduce the Adaptive Cross-Hadamard (ACH) module, a novel operator that embeds learnability through differentiable discrete sampling and dynamic softsign normalization. This facilitates highly efficient feature reuse without incurring additional convolutional parameters, while ensuring stable gradient flow. Integrated into Hadaptive-Net (Hadamard Adaptive Network) via neural architecture search, our approach achieves unprecedented efficiency. Comprehensive experiments demonstrate state-of-the-art accuracy/speed trade-offs on image classification tasks, establishing Hadamard operations as specific building blocks for efficient vision models.