TY  - RPRT
TI  - CapFrame: Text-Instructed Viewpoint Grounding in 3D Gaussian Scenes via Geometric Pseudo Labels
AU  - Jirong Li
AU  - Satoshi Ikehata
AU  - Shuhei Kurita
AU  - Ikuro Sato
PY  - 2026
UR  - https://arxiv.org/abs/2608.30342
ID  - 2608.30342
ER  -