Onur Keleş

Onur Keleş

PhD researcher at Boğaziçi University and IMPRS Affiliate at Max Planck Institute for Psycholinguistics, working on multimodality, iconicity, sign languages, and computational linguistics.

The Visual Iconicity Challenge: Evaluating Vision–Language Models on Sign Language Form–Meaning Mapping

Published in Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2026, Main), 2026

We introduce the Visual Iconicity Challenge, a benchmark for evaluating whether vision–language models can map iconic form–meaning relationships in sign languages. Our experiments reveal significant limitations in current VLMs in understanding visual iconicity.

Preprint on arXiv

Recommended citation: Keleş, O., Özyürek, A., Ortega, G., Gökgöz, K., & Ghaleb, E. (2025). The Visual Iconicity Challenge: Evaluating Vision–Language Models on Sign Language Form–Meaning Mapping. arXiv:2510.08482. Accepted at ACL 2025 (Main conference).