ChiMed-VL: 580,014 Chinese Medical Image-Text Pairs for Vision-Language Alignment

ChiMed-VL-Alignment is a multimodal dataset containing 580,014 Chinese medical image-text pairs. The dataset was created by author 'williamliu' and was last updated on the Hugging Face platform in December 2023. Pairs are categorized into context information and image-specific descriptions, with the context category containing 167 million tokens and descriptions containing 63 million tokens.

Use Cases

• Training vision-language models for medical image captioning based on the dataset's descriptive image-text pairs.
• Fine-tuning models for medical visual question answering using the context information provided with images.
• Benchmarking model performance on Chinese medical image-text alignment tasks.
• Developing multimodal retrieval systems for medical images based on textual queries.

Strengths

• Contains 580,014 image-text pairs, providing a substantial volume of training data.
• Offers two distinct categories of text: context information (167M tokens) and descriptions (63M tokens).
• Provides detailed token statistics, such as a median text length of 435 tokens for context information.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last updated 2023-12-01 12:37:17; freshness should be verified.

Provenance

Source

huggingface

Freshness

2023-12-01 12:37:17

Geography

Likely China or Chinese-language sources, inferred from the dataset's name and focus.