In six VQA benchmarks, models can often choose the correct option from image plus options alone, and the GroundAttack toolkit generates visually plausible hard negatives to remove this shortcut.
Vizwiz grand challenge: Answering visual questions from blind people
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
dataset 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
dataset 1polarities
background 1representative citing papers
citing papers explorer
-
Mitigating Easy Option Bias in Multiple-Choice Question Answering
In six VQA benchmarks, models can often choose the correct option from image plus options alone, and the GroundAttack toolkit generates visually plausible hard negatives to remove this shortcut.