EgoP$^{2}$C: Internalizing Plan-Perceive-Compose for Spatial Reasoning in Egocentric Video
Answering complex spatial reasoning questions from egocentric video requires determining what evidence is needed, acquiring targeted observations, and combining them through multiple reasoning steps. Existing vision-language models (VLMs) struggle to perform this process reliably.
arXiv - cs.CV (CV)2 min read