VisAffect at MWE-2026 AdMIRe 2: IMMCAN Idiom Multimodal Cross-Attention Network

Research summary

Studies multilingual image ranking using frozen language and vision–language encoders with a trainable cross-attention module. Visual grounding improves zero-shot transfer, while text-only modeling performs better on the smaller in-domain test.

DOI: 10.18653/v1/2026.mwe-1.19

Citation

Barış Bilen, Ali Azmoudeh, Hazım Kemal Ekenel, Hatice Köse. 2026. VisAffect at MWE-2026 AdMIRe 2: IMMCAN Idiom Multimodal Cross-Attention Network. Proceedings of the 22nd Workshop on Multiword Expressions (MWE 2026).

@inproceedings{bilen2026visaffect,
  title = {{VisAffect at MWE-2026 AdMIRe 2: IMMCAN Idiom Multimodal Cross-Attention Network}},
  author = {Barış Bilen and Ali Azmoudeh and Hazım Kemal Ekenel and Hatice Köse},
  year = {2026},
  booktitle = {Proceedings of the 22nd Workshop on Multiword Expressions (MWE 2026)},
  doi = {10.18653/v1/2026.mwe-1.19},
  url = {https://aclanthology.org/2026.mwe-1.19/},
  pages = {149--153},
  publisher = {Association for Computational Linguistics}
}