TY  - RPRT
TI  - OVSeg3R: Learn Open-vocabulary Instance Segmentation from 2D via 3D Reconstruction
AU  - Hongyang Li
AU  - Jinyuan Qu
AU  - Lei Zhang
PY  - 2026
UR  - https://arxiv.org/abs/2509.23541
ID  - 2509.23541
ER  -