3DSRBench

The '3DSRBench' is a benchmark dataset used to evaluate the performance of multimodal language models on visual perspective-taking tasks, particularly in assessing their ability to adopt different spatial orientations.

Papers using 3DSRBench (3)

Enhancing MLLM Spatial Understanding via Active 3D Scene Exploration for Multi-Perspective Reasoning2026SpatiO: Adaptive Test-Time Orchestration of Vision-Language Agents for Spatial Reasoning2026Cognitively-Inspired Tokens Overcome Egocentric Bias in Multimodal Models2026