CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models
Cross-image comparative reasoning remains challenging for vision-language models (VLMs), especially when correct prediction requires fine-grained attribute grounding and globally consistent reasoning. We present CoRe, a unified framework for this problem. CoRe includes: (i) CoRe-20K, a large-scale triplet-based training set automatically constructed from structured visual metadata through a multi-expert collaborative pipeline, covering counting, depth, distance, and spatial relations; (ii) TriSR, a structured reward framework that jointly supervises attribute grounding, judgment alignment, and
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- FuzzySimilar title/name (fuzzy) · 59%Tongyi-MAI/Z-Image-Turbo →
“Fuzzy title match (0.73): “CoRe: A Comprehensive Framework for Cross-Image Comparative ” ≈ “Tongyi-MAI/Z-Image-Turbo””
- FuzzySimilar title/name (fuzzy) · 59%VioletVision-3B →
“Fuzzy title match (0.73): “CoRe: A Comprehensive Framework for Cross-Image Comparative ” ≈ “VioletVision-3B””
- LinkedLinked via arxiv author · 85%Lin Peng →
“CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models”
- LinkedLinked via arxiv author · 85%Cong Wang →
“CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models”
- LinkedLinked via arxiv author · 85%Zeyu Guo →
“CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models”
- LinkedLinked via arxiv author · 85%SongLin Dong →
“CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models”
- LinkedLinked via arxiv author · 85%Yihong Gong →
“CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models”
- FuzzySimilar title/name (fuzzy) · 84%pytorch/vision →
“Fuzzy title match (0.92): “CoRe: A Comprehensive Framework for Cross-Image Comparative ” ≈ “pytorch/vision””