#machine learning
Jan 2026
SpaRRTa: A Synthetic Benchmark for Evaluating Spatial Intelligence in Visual Foundation Models
The Spatial Relation Recognition Task (SpaRRTa) benchmark is introduced, which evaluates the ability of VFMs to identify relative positions of objects in the image and provides insights into the mechanisms that support or hinder spatial awareness in modern VFMs.
Turhan Can Kargin, Wojciech Jasinski, Adam Pardyl et al.
· arXiv.org · 0 citations