AI-Assisted Painting Restoration User Study Data

34 participants provided experimental data for a study on AI-assisted restoration of ancient Chinese paintings. The dataset includes training data for stylized LoRA models and results from AI and non-AI user groups. Liu Zhongni published this replication data in March 2026.

Use Cases

• Compare restoration outcomes between AI-assisted participants (A1-A17) and non-AI participants (N1-N17).
• Fine-tune stylized LoRA models using the provided training datasets.
• Analyze participant consent and study design using the included informed consent form.

Strengths

• 34 total participants across two experimental groups
• Includes original training data for model development
• Replication data for a published 2026 manuscript

Limitations

• Unknown sample size for the training datasets
• User study limited to 34 participants
• Geographic and cultural scope of participants is unspecified

Provenance

Source

Liu, Zhongni via Humanities & Social Sciences Communications Dataverse

Collection Method

Experimental data collected from a user study.

Freshness

Last updated March 2026