SemEval-2025 Task 1: Advancing Multimodal Idiomaticity Representation

Join the SemEval-2025 Shared Task on Idiomaticity Detection in Images and help advance multimodal understanding of idiomatic expressions in visual contexts. Improve the performance of applications such as sentiment analysis, machine translation, and natural language understanding.

We invite you to participate in the SemEval-2025 Shared Task on Idiomaticity Detection in Images, a novel challenge focused on advancing multimodal understanding through the analysis of idiomatic expressions in visual contexts.

Comparing the performance of language models to humans shows that models lag behind humans in comprehension of idioms. To address this, we build on the previous SemEval-2022 Task 2 and seek to explore the comprehension ability of multimodal models that incorporate visual and textual information. The goal is to understand how well these models can capture representations and whether multiple modalities can improve these representations.

Good representations of idioms are crucial for applications such as sentiment analysis, machine translation, and natural language understanding. Exploring ways to improve models’ ability to interpret idiomatic expressions can enhance the performance of these applications.

Tags: SemEval-2025, Task 1, Idiomaticity Detection, Multimodal Understanding, Natural Language Processing, Idioms, Visual Context, Language Models