ArabicNLP2026-ImageEval 2026

Deadlines
Computer Vision/CORE Unranked

ArabicNLP2026-ImageEval 2026

ArabicNLP 2026 Shared Task on ImageEval 2026: Culturally Grounded Arabic Multimodal Evaluation

1792789200000Budapest, HungaryOfficial conference site Site reachable

ImageEval 2026 is a shared task focused on culturally grounded Arabic multimodal AI, evaluating visual question answering, image-grounded hallucination detection, and cultural accuracy in text-to-image generation. It emphasizes Arabic language varieties and culturally specific visual content from the MENA region, with two main tasks: AynVQA and CRAI-Bench.

Paper fit

Contribution paths

A strong submission should clearly identify its contribution and evaluate it appropriately.

System Description Paper

A paper explaining the team's system, training, evaluation, and analysis, limited to 4 pages for submission and 5 pages for camera-ready, with unlimited references and appendix. Must follow EMNLP 2026 / ACL style templates and include author names and affiliations.

Research areas in scope

01

Tasks

AynVQA - Culturally Grounded Arabic Visual Question Answering and Hallucination DetectionCRAI-Bench - Cultural Accuracy Evaluation for Text-to-Image Generation

Policies worth checking twice

  • Papers are not anonymized; author names, affiliations, and contact information must be included.
  • Maximum 4 content pages for initial submission; camera-ready papers may use up to 5 content pages.
  • Unlimited pages for references and appendix.
  • Papers must use the official EMNLP 2026 / ACL LaTeX or Word templates; non-conforming formatting may be rejected.
  • All external data, tools, APIs, or models used beyond the released task data must be documented.
  • Official submitted results and post-submission experiments must be clearly distinguished.
  • Camera-ready papers must pass ACL PubCheck before submission.
  • System description papers must cite the ImageEval 2026 overview paper and relevant dataset papers.

Official sources

Compiled from the official call for papers. The organizers’ pages remain authoritative.

Last verified September 9, 2026