Visual Question-Answering for Thematic Maps (NFDI4Earth Incubator Lab 2025) (VISQAM)
Basic data for this project
Type of project: Subproject in DFG-joint project hosted outside University of Münster
Duration at the University of Münster: 01/06/2025 - 31/12/2025 | 1st Funding period
Description
Visual question-answering (QA) helps users interpret complex visual information, making it easier and faster to gain insights also from maps and geospatial data in a variety of contexts. This project aims to create an open dataset tailored for thematic map-based QA systems, accompanied by a baseline model to demonstrate its usage. By compiling map images annotated with question-answer pairs, the dataset will enable Artificial Intelligence (AI) models to extract and interpret geographic and information from maps. The deliverables will include a curated dataset, a baseline model, documentation, and an evaluation report, all of which will be released under a permissive license to support further research on the topic.
Keywords: Open data; Visual Question-Answering; VQA; Large Language Models
Website of the project: https://www.nfdi4earth.de/2participate/incubator-lab
Funder / funding scheme:
- DFG - National Research Data Infrastructure (NFDI)
Project management at the University of Münster
| Koukouraki, Eftychia |
Applicants from the University of Münster
| Koukouraki, Eftychia |
Coordinating organisations outside the University of Münster
- Dresden University of Technology (TU Dresden)Germany