Semantic Segmentation–Guided Reconstruction and Artistic Style Synthesis of Intangible Cultural Heritage Patterns Using a Deep Learning Framework

This dataset contains 15,148 images of traditional Miao Batik cultural patterns used for semantic segmentation, pattern reconstruction, and artistic style generation tasks. The dataset includes annotated image–mask pairs representing diverse motif categories such as animal, plant, and geometric designs. All images are preprocessed to a uniform resolution of 256 × 256 pixels, and corresponding segmentation masks are provided to facilitate supervised learning. The dataset is organized with associated metadata, including image paths, mask paths, and categorical labels, enabling reproducible experimentation. It was used in conjunction with a deep learning framework integrating SegFormer-B2 for segmentation, CycleGAN for pattern reconstruction, and SPADE for style synthesis. This dataset supports research in cultural heritage preservation, computer vision, and generative modeling, and is made publicly available to promote transparency and reproducibility.