ViInfographicsVQA
A Vietnamese infographic question-answering project that follows the path from visual data to a working web demo.
The problem
An infographic combines text, layout, and visual information. Answering a question requires understanding them together.
The approach
The team collected images, generated question–answer pairs, analyzed the data, trained models, and evaluated generated answers.
Available materials
The repository links the dataset and models, with Streamlit / FastAPI demos that make the workflow easier to inspect.
Contribution & context
Team contributor for Text Mining and Applications, February–May 2025.
Scope of this project summary
This is a shared team project. The page does not assign individual ownership or claim a published benchmark result.