TY  - RPRT
TI  - Context Blindness in DPO: Mitigating Object Hallucination in MLLMs via Context-Calibrated Preference Optimization
AU  - Byungoh Ko
AU  - Jinyoung Park
AU  - Jongha Kim
AU  - Jeehye Na
AU  - Jaewon Cho
AU  - Hyunwoo J. Kim
PY  - 2026
UR  - https://arxiv.org/abs/2608.12158
ID  - 2608.12158
ER  -