{"@context":"https://schema.org","@type":"CreativeWork","@id":"https://forgecascade.org/public/capsules/fd31e8fe-f442-41aa-9655-f06320c7c5a9","identifier":"fd31e8fe-f442-41aa-9655-f06320c7c5a9","url":"https://forgecascade.org/public/capsules/fd31e8fe-f442-41aa-9655-f06320c7c5a9","name":"WikiVQABench: A Knowledge-Grounded Visual Question Answering Benchmark from Wikipedia and Wikidata","text":"# WikiVQABench: A Knowledge-Grounded Visual Question Answering Benchmark from Wikipedia and Wikidata\n\n**Authors:** Basel Shbita, Pengyuan Li, Anna Lisa Gentile\n**arXiv:** https://arxiv.org/abs/2605.21479v1\n**Published:** 2026-05-20T17:58:24Z\n\n## Abstract\nVisual Question Answering (VQA) benchmarks have largely emphasized perception-based tasks that can be solved from visual content alone. In contrast, many real-world scenarios require external knowledge that is not directly observable in the image to answer correctly. We introduce WikiVQABench, a human-curated knowledge-grounded VQA benchmark constructed by systematically combining Wikipedia images, their associated article captions, and structured knowledge from Wikidata. Our pipeline uses large language models (LLMs) to generate candidate multiple-choice image-question-answer sets. All generated instances are subsequently reviewed and curated by human annotators to ensure factual correctness, visual-text consistency, and that each question requires external knowledge in addition to visual evidence for correct resolution. WikiVQABench comprises a substantial collection of Wikipedia images with curated multiple-choice questions designed to benchmark knowledge-aware vision-language models (VLMs). Evaluation of fifteen VLMs (256M-90B parameters) reveals a wide performance range (24.7%-75.6% accuracy), demonstrating that the benchmark effectively discriminates model capabilities on knowledge-intensive reasoning. The dataset and benchmarking code are publicly available.","keywords":["cs.CV","cs.AI"],"about":[{"@type":"Thing","name":"ifconfig"},{"@type":"Thing","name":"pwdump"},{"@type":"Thing","name":"Standard Encoding"},{"@type":"Thing","name":"Botnet"},{"@type":"Thing","name":"Artificial Intelligence"}],"citation":[],"isPartOf":{"@type":"Dataset","name":"Forge Cascade Knowledge Graph","url":"https://forgecascade.org"},"publisher":{"@type":"Organization","name":"Forge Cascade","url":"https://forgecascade.org"},"dateCreated":"2026-05-21T06:00:06.267000Z","dateModified":"2026-05-21T06:00:06.267000Z","isBasedOn":"https://arxiv.org/abs/2605.21479v1","additionalProperty":[{"@type":"PropertyValue","name":"trust_level","value":65},{"@type":"PropertyValue","name":"verification_status","value":"source_linked"},{"@type":"PropertyValue","name":"evidence_level","value":"primary_source"}]}