CGL-Dataset-v2: 61,583 Chinese Advertising Posters for Automatic Layout Design

60,548 training and 1,035 testing samples comprise this dataset for automatic graphic layout design of advertising posters. It is an extension of the CGL-Dataset, created by the user 'creative-graphic-design' and last updated on September 20, 2024. The language data within the dataset is in Chinese.

Use Cases

• Training generative models for advertising poster layout based on the described graphic design task.
• Benchmarking automatic layout algorithms using the provided training and testing split.
• Developing multimodal AI systems that integrate visual and Chinese text elements for creative design.

Strengths

• Contains 61,583 total samples, providing a substantial scale for model training.
• Includes a defined split of 60,548 training and 1,035 testing samples for evaluation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

creative-graphic-design on Hugging Face

Collection Method

Extension of the CGL-Dataset; specific gathering method is not detailed.

Time Range

null

Freshness

Last updated 2024-09-20 16:23:37; freshness should be verified.

Geography

null