Lightweight Artistic Style Classification via Texture Correlation and Adaptive Local–Global Fusion

Abstract

Artistic style classification remains challenging because visually similar styles often exhibit subtle differences in brushstrokes, textures, colors, and global composition. To address this problem, this paper proposes a lightweight artistic style classification framework based on a MobileNet–Transformer Hybrid (MTH). First, a Lightweight Texture Correlation Module (LTCM) is introduced to capture multi-scale texture relationships from intermediate convolutional features. Second, an Adaptive Local–Global–Texture Fusion module (ALGTF) dynamically integrates local details, global composition, and texture representations according to their relative contributions. Third, a Confusable-Style Hard Contrastive Learning strategy (CSHCL) is designed to emphasize difficult inter-style samples and improve feature separability between visually similar artistic categories. The Dynamic Channel-Spatial Attention Module (DCSAM) is retained to enhance style-sensitive regions and channels. Experiments on ArtBench-10 and Painting91_Styles demonstrate that the proposed framework provides improved classification performance while maintaining a lightweight architecture, showing its effectiveness for fine-grained artistic image analysis.