Introduces the first benchmark for metaphorical video understanding, identifies MLLM weaknesses in cross-domain mapping, and proposes an inference-time enhancement using a knowledge graph.
Vrr-qa: Visual relational rea- soning in videos beyond explicit cues.arXiv preprint arXiv:2506.21742, 2026
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
fields
cs.CV 5years
2026 5verdicts
UNVERDICTED 5representative citing papers
A pipeline using question-aware evidence ledgers with GPT-5.5 achieves 92.95% overall and 93.79% macro accuracy on the VRR-QA video relational reasoning challenge.
Empirical study on ImplicitQA benchmark finds video QA is perception-bound, with reasoning augmentations neutral or harmful and low-level perception categories hardest.
ASC-MQRA applies answer self-consistency across stochastic video QA runs and optional margin-triggered re-arbitration to achieve 81.16% average accuracy on the CVPR 2026 VidLLMs Challenge Track 2 test set.
An adaptive test-time system for VRR-QA challenge achieves 90.07 average accuracy by routing difficult questions to a dense evidence refinement module after initial model passes.
citing papers explorer
-
MetaphorVU: Towards Metaphorical Video Understanding
Introduces the first benchmark for metaphorical video understanding, identifies MLLM weaknesses in cross-domain mapping, and proposes an inference-time enhancement using a knowledge graph.
-
Question-Aware Evidence Ledgers for Video Relational Reasoning
A pipeline using question-aware evidence ledgers with GPT-5.5 achieves 92.95% overall and 93.79% macro accuracy on the VRR-QA video relational reasoning challenge.
-
Perception First: A Frontier Native-Video Model with Self-Consistency for Implicit Video Question Answering
Empirical study on ImplicitQA benchmark finds video QA is perception-bound, with reasoning augmentations neutral or harmful and low-level perception categories hardest.
-
Answer Self-Consistency with Margin-Triggered Question Re-Arbitration for the CVPR 2026 VidLLMs Challenge
ASC-MQRA applies answer self-consistency across stochastic video QA runs and optional margin-triggered re-arbitration to achieve 81.16% average accuracy on the CVPR 2026 VidLLMs Challenge Track 2 test set.
-
Adaptive Dense Evidence Refinement for Video Relational Reasoning for VRR-QA Challenge
An adaptive test-time system for VRR-QA challenge achieves 90.07 average accuracy by routing difficult questions to a dense evidence refinement module after initial model passes.