Multi-view relational distillation transfers geometric knowledge to vision-language models by matching cross-view patch similarity matrices, improving spatial reasoning with minimal overhead.
Henry Holt and Co., Inc., New York, NY, USA
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Multi-View Relational Distillation for Spatial Reasoning with Vision-Language Models
Multi-view relational distillation transfers geometric knowledge to vision-language models by matching cross-view patch similarity matrices, improving spatial reasoning with minimal overhead.