Dear all,

We are offering a task on Multimodal Understanding of Smells in Texts and Images (MUSTI) within the MediaEval benchmark on the detection of smell-oriented relations in historical images and historical texts in English, French, German, and Italian. Our focus is on determining whether an image and text refer to the same smell source(s) and what are the smell sources that make this pair related. The data sets contain several thousand pairs of images and texts annotated by native speakers who are experts in olfactory information processing.

This is the second iteration of MUSTI. We are adding a zero-shot task this year. Please refer to the following publications for an overview [1] and baselines [2] for more information.

Registration to participate is now open, and submissions will be due in late November (the deadline for submission to be adjusted). You will have the chance to present your work at the 14th Annual MediaEval Workshop 1-2 February 2024, which is collocated with MMM 2024 in Amsterdam, the Netherlands, with remote participation also possible.

More information can be found on

MUSTI: https://multimediaeval.github.io/editions/2023/tasks/musti/

MediaEval: https://multimediaeval.github.io/editions/2023/

MMM: https://mmm2024.org

We would like to invite you to participate in this task. Please let us know if you have any questions. We are looking forward to hearing from you!

Best regards,

Ali Hürriyetoğlu (on behalf of organizing committee)

[1] Hürriyetoğlu, A., Paccosi, T., Menini, S., Mathias, Z., Pasquale, L., Kiymet, A., ... & van Erp, M. (2022). MUSTI-Multimodal Understanding of Smells in Texts and Images at MediaEval 2022. In Proceedings of MediaEval 2022 CEUR Workshop. URL: https://cris.fbk.eu/retrieve/a8af3795-9055-4cee-8465-f4194a05c5d5/paper9634....

[2] Akdemir, K., Hürriyetoğlu, A., Troncy, R., Paccosi, T., Menini, S., Zinnen, M., Christlein, V. (2022). Multimodal and Multilingual Understanding of Smells using VilBERT and mUNITER. In Proceedings of MediaEval 2022 CEUR Workshop. URL: https://cris.fbk.eu/retrieve/6112053b-b228-41e9-b763-c33acc876da9/paper6505....