2026-10-11 · America/Los_Angeles · 论文 · #13

Benchmarking Answer-Grounding Consistency in Remote Sensing Visual Question Answering

In grounded remote sensing visual question answering, a model produces a textual answer and identifies relevant objects in the image as grounding evidence. Existing evaluations measure answer correctness and localization quality, but do not directly assess answer-grounding consistency. We propose the Answer-Evidence Support Protocol (AESP) to assess this consistency through evidence support: whether the grounding evidence supports the model's own answer. AESP applies task-specific rules to existence, counting, and comparison questions. We further construct the Grounded Remote Sensing Visual…

热度 15.3 / 100；排名与评分保留该期记录。