2 papers
cs.AI2026
OrientSAM: Mitigating Camera-Centric Shortcut in Multimodal Spatial Reasoning via Orientation-Aware Spatial Alignment
Wenxiao Fan, Hang Yin, Kan Li
Multimodal large language models (MLLMs) still struggle with spatial reasoning that requires perspective transformation. In particular, they often rely on camera-centric cues rathe…
cs.AI2024
Do Multimodal Language Models Really Understand Direction? A Benchmark for Compass Direction Reasoning
Hang Yin, Zhifeng Lin, Xin Liu +2
Direction reasoning is essential for intelligent systems to understand the real world. While existing work focuses primarily on spatial reasoning, compass direction reasoning remai…