TY  - RPRT
TI  - Boosting Visual Knowledge-Intensive Training for LVLMs Through Causality-Driven Visual Object Completion
AU  - Qingguo Hu
AU  - Ante Wang
AU  - Jia Song
AU  - Delai Qiu
AU  - Qingsong Liu
AU  - Jinsong Su
PY  - 2025
UR  - https://arxiv.org/abs/2508.04453
ID  - 2508.04453
ER  -