DIQ-H is a new benchmark for VLM performance under continuous adversarial image degradations, paired with VIR that raises annotation accuracy from 72.2% to 83.3%.
Neg- ative Object Presence Evaluation (NOPE) to Measure Object Hallucination in Vision-Language Models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Value-Guided Iterative Refinement and the DIQ-H Benchmark for Evaluating VLM Robustness
DIQ-H is a new benchmark for VLM performance under continuous adversarial image degradations, paired with VIR that raises annotation accuracy from 72.2% to 83.3%.