ViTextVQA: 50,000 Vietnamese Text-Based Visual Question Answer Pairs

ViTextVQA contains over 16,000 images and 50,000 question-answer pairs focused on Vietnamese text comprehension within visual contexts. Developed by researcher minhquan6203 and documented in Arxiv paper 2404.10652, it serves as a benchmark for text-based visual question answering in the Vietnamese language.

Use Cases

• Training models to extract and interpret Vietnamese text from images using the question-answer pairs
• Evaluating OCR-based visual question answering systems on Vietnamese linguistic nuances
• Benchmarking multimodal LLMs on regional language text comprehension

Strengths

• Contains over 50,000 question-answer pairs
• Includes 16,000+ images with Vietnamese text
• First large-scale dataset specifically for Vietnamese text-based VQA

Limitations

• Specific column names and schema are not documented in the provided metadata
• Restricted to non-commercial use under CC BY-NC 3.0 license

Provenance

Source

minhquan6203 (Arxiv:2404.10652)

Collection Method

Annotated

Freshness

Last updated January 2026.

Geography

Vietnam