An agent that cross-validates a VideoLLM's answer with YOLO-World open-vocabulary grounded object tracking sets new zero-shot state-of-the-art scores on three VideoQA benchmarks.
Gemini 1.5: Unlocking multimodal un- derstanding across millions of tokens of context, 2024
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
-
ViQAgent: Zero-Shot Video Question Answering via Agent with Open-Vocabulary Grounding Validation
An agent that cross-validates a VideoLLM's answer with YOLO-World open-vocabulary grounded object tracking sets new zero-shot state-of-the-art scores on three VideoQA benchmarks.