VIKSER combines fine-grained visual relationship captions, question paraphrasing, evidence-based prompting, and self-reflection to reach state-of-the-art results on six visual reasoning datasets.
Interpretable visual reasoning: A survey,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
VIKSER: Visual Knowledge-Driven Self-Reinforcing Reasoning Framework
VIKSER combines fine-grained visual relationship captions, question paraphrasing, evidence-based prompting, and self-reflection to reach state-of-the-art results on six visual reasoning datasets.