一作ACL 2026WorkshopBest SRW Paper· Natural Language Processing
Evaluation of Multilingual Ability to Use Spatial Deictic Expressions in Vision-Language Models
Kaito Watanabe · Taisei Yamamoto · Tomoki Doi · Hitomi Yanaka
摘要
One of the expected abilities of vision-language models (VLMs) is spatial reasoning ability based on a given text and image.To evaluate the spatial reasoning abilities of VLMs, we focus on the use of spatial deictic expressions, which are defined as spatial expressions whose referent is determine...