Zhuxian New Year Pictures Dataset for AI-Generated Art Re-Creation

A high-quality dataset built through a three-stage process of image acquisition, multidimensional screening, and expert review for the digital re-creation of traditional Chinese folk art. The dataset supports a technical framework combining the Liblib platform and a LoRA model to explore AI-generated art. It was created by Jiangxu Zhang and last updated on 2026-05-27.

Use Cases

• Training AI models for style transfer based on the stylistic features of Zhuxian New Year pictures mentioned in the description
• Developing cultural and graphic creative products based on the integration of traditional aesthetics with modern design
• Researching digital inheritance and modernization methodologies for folk arts based on the described collaborative technical framework
• Evaluating image generation quality based on the expert-reviewed, multi-stage screening process used to create the dataset

Strengths

• Dataset construction involved a three-stage process including expert review, suggesting quality control.
• A three-layer keyword thesaurus was constructed through literature analysis, questionnaire surveys, and semantic clustering.
• The associated model uses a two-stage training strategy with dynamic optimization to capture artistic style.

Limitations

• Dataset size is only 51.7 KB, indicating a very limited scope.
• Row count and column-level documentation are unknown, limiting suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare, author Jiangxu Zhang

Collection Method

Built through image acquisition, multidimensional screening, and expert review.

Freshness

Last updated 2026-05-27 17:40:18; freshness should be verified.

Geography

Focuses on Zhuxian Town, China.